Skip to main content

Full text of "Proceedings : International Workshop on Subalpine Stone Pines and Their Environment: the Status of Our Knowledge, St. Moritz, Switzerland, September 5-11, 1992"

See other formats


Historic,  Archive  Document 

Do  not  assume  content  reflects  current 
scientific  knowledge,  policies,  or  practices. 


United^tes 
Department 
Agriculture 


Forest  Service 

Intermountain 
Research  Station 

General  Technical 
Report  INT-GTR-309 


July  1994 


Proceedings — International 
Workshop  on  Subalpine 
Stone  Pines  and  Their 
Environment:  the  Status 
of  Our  Knowledge 


n 


:3 


Statements  by  contributors  from  outside  the  U.S.  Department 
of  Agriculture  are  solely  the  views  of  the  authors  and  do  not  re- 
flect policies  or  opinions  of  the  Department. 

The  use  of  trade  or  firm  names  in  this  publication  is  for  reader 
information  and  does  not  imply  endorsement  of  any  product  or 
service  by  the  U.S.  Department  of  Agriculture  or  other  organi- 
zations represented  here. 


Intermountain  Research  Station 
324  25th  Street 
Ogden,  UT  84401 


Proceedings — International 
Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment: 

the  Status  of  Our  Knowledge 

St.  Moritz,  Switzerland,  September  5-11, 1992 


Compilers: 

Wyman  C.  Schmidt 
Intermountain  Research  Station 
Forest  Service 

U.S.  Department  of  Agriculture 

Friedrich-Karl  Holtmeier 
Department  of  Geography 
University  of  Munster 
Germany 


Workshop  Sponsors: 

intermountain  Research  Station 
Westfalische  Wilhelms-Universitat, 

Landscape  Ecology  Division,  Germany 
Swiss  Federal  Institute  for  Forest,  Snow, 

and  Landscape  Research 
Karl  and  Sophie  Binding  Foundation, 

Switzerland 


FOREWORD 

This  proceedings  is  a  product  of  the  first  comprehensive 
and  international  examination  of  stone  pine  species  of  the 
world.  It  reports  the  papers  that  were  presented  at  the  Inter- 
national Workshop  on  "Subalpine  Stone  Pines  and  Their 
Environment — The  Status  of  Our  Knowledge"  held 
September  5-12, 1992,  at  St.  Moritz,  Switzerland. 

It  is  generally  recognized  that  there  are  five  subalpine 
stone  pines  in  the  world— all  in  the  Northern  Hemisphere- 
one  in  North  America  and  four  in  Europe  and  Asia.  Ail  share 
many  of  the  same  characteristics  and  occupy  similar  ecologi- 
cal niches.  Stone  pines  are  defined  as  those  species  of  Pinus 
that  have  five-needled  fascicles,  have  wingless  seeds,  and 
have  cones  that  remain  closed  at  maturity. 

How  did  this  international  workshop  come  about?  Comple- 
tion of  a  whitebark  pine  workshop  in  1987  and  a  major  sym- 
posium in  1989  in  Bozeman,  MT,  U.S.A.,  sparked  interest  in 
high-elevation  ecosystems  of  western  North  America.  One 
section  of  this  popular  symposium  specifically  addressed  high- 
elevation  ecosystems  of  the  world.  Information  presented 
there  hinted  that  we  could  learn  much  about  whitebark  pine 
ecosystem  analogs  from  other  parts  of  the  world  that  have 
similar  physical  and  biological  features.  In  addition  many  of 
the  Eurasian  forests  have  a  long  history  of  human  influence. 
The  logical  place  to  look  for  this  type  of  information  was  the 
world's  forest  areas  that  supported  subalpine  stone  pines. 
A  cursory  examination  of  the  five  commonly  recognized  sub- 
alpine stone  pines  indicated  they  had  many  similarities. 

With  this  as  a  background,  a  small  group  of  western  U.S. 
researchers  and  managers  held  a  planning  session  with 
Dr.  Holtmeier  from  Muenster,  Germany,  in  Missoula,  MT,  in 
early  1990.  This  planning  spawned  the  notion  that  it  was 
time  that  we  started  doing  a  better  job  of  collaborating  be- 
tween colleagues  in  Europe,  Asia,  and  North  America.  The 
logical  starting  point  was  an  international  workshop  that  ad- 
dressed those  items  that  held  common  interest  for  European, 
Asian,  and  North  American  specialists.  So,  the  international 
planning  transmission  was  put  into  gear  and  2V2  years  later  a 
workshop  was  held  in  St.  Moritz,  Switzerland. 

The  primary  objectives  of  the  Stone  Pine  Workshop  were: 

1 .  To  improve  international  cooperation  and  collaboration 
between  scientists. 

2.  To  exchange  research  findings. 

3.  To  determine  knowledge  gaps  in  the  ecology  and  man- 
agement of  subalpine  stone  pines  of  the  world. 

All  five  subalpine  stone  pines  were  covered  by  authors  of 
48  papers  from  Europe,  Asia,  and  North  America.  They  were: 


Species 

Swiss  stone  pine 
{Pinus  cembra) 

Siberian  stone  pine 
{Pinus  sibirica) 


Natural  range 

European  Alps,  Carpathian 
Mountains 


Siberian  Russia  and  Northern 
Mongolia  (95  percent  in 
Russia,  5  percent  in  Mongolia) 

Japanese  stone  pine     Japan,  Korea,  Eastern  Siberia 
{Pinus  pumila) 

Korean  stone  pine        Northeastern  China,  North  Korea, 
{Pinus  koraiensis)       Honshu  Japan,  southeastern 

Russia 


Whitebark  pine 
{Pinus  albicaulis) 


Western  North  America 


The  1 2  countries  represented  and  the  number  of  papers 
from  each  country  were:  Russia  (9),  China  (1),  Japan 
(4),  Switzerland  (5),  Austria  (4),  Italy  (3),  France  (1),  Yugosla- 
via (1),  Czechoslovakia  (1),  Romania  (1),  Germany  (3),  and 
United  States  (15).  The  main  subjects  covered  were: 

•  Evolution  and  taxonomy. 

•  Basic  ecology  of  stone  pine  species. 

•  Growth  characteristics. 

•  Influence  of  environmental  factors. 

•  Regeneration. 

•  Importance  to  wildlife. 

•  Forest  structure  and  dynamics. 

•  Forest  management. 

•  Research  needs. 

Stone  pines  are  critical  forest  components  on  vast  areas. 
Pinus  sibirica  alone  covers  40  million  hectares,  mainly  in 
Siberia,  and  is  a  key  component  of  the  economy  and  biology. 
P.  l<oraiensis  accounts  for  one-fourth  of  the  timber  production 
in  China.  P.  albicaulis  is  key  to  the  survival  of  wildlife  such  as 
the  grizzly  bear  and  is  a  protector  of  high-elevation  water- 
sheds and  visual  resources.  P.  cembra  plays  a  key  role  in 
avalanche  protection,  wood  production,  and  recreation  in 
Europe.  P.  pumila  is  a  key  high-elevation  species  for  stabili- 
zation of  mountain  slopes  in  Japan  and  Kamchatka.  These 
species  have  many  things  in  common,  likely  starting  with  a 
common  ancestor,  probably  P.  sibirica.  They  also  share 
common  methods  of  seed  dispersal  with  the  North  American 
and  Eurasian  nutcrackers.  Natural  selection  processes  have 
resulted  in  a  real  mutualism  between  subalpine  stone  pines 
and  the  nutcrackers. 

This  workshop  was  sponsored  by  the  Intermountain 
Research  Station,  USDA  Forest  Service,  U.S.A.;  the 
Westfalische  Wilhelms-Universitat,  Landscape  Ecology  Divi- 
sion, Germany;  the  Swiss  Federal  Institute  for  Forest,  Snow, 
and  Landscape  Research,  Switzerland;  and  the  Karl  and 
Sophie  Binding  Foundation,  Switzerland. 

We  want  to  acknowledge  the  people  and  the  organizations 
that  supported  them  for  their  enthusiastic  and  dedicated  ef- 
forts in  the  planning  and  conduct  of  this  international  work- 
shop. Three  field  tours  were  conducted  by  the  hosts  of  the 
workshop.  The  first  tour  entitled  "Engadine  Cembra  Pine- 
Larch  Forest,"  was  in  a  cembra  pine-larch  forest  overlooking 
St.  Moritz,  a  heavily  used  recreation  area  in  the  Engadine. 
The  primary  subjects  were  tree  regeneration  and  succession 
in  avalanche  tracks  and,  at  the  upper  timberline,  squirrel  and 
nutcracker  interactions,  human  historical  influence,  grazing 
and  game,  and  the  pressures  of  modern  tourism  on  the  for- 
ests and  landscape. 

The  second  tour,  entitled  "Ecology  and  Technique  of  High- 
Altitude  Afforestation,"  dealt  primarily  with  management  of 
avalanche-prone  areas.  It  was  conducted  in  the  Stillberg  Ex- 
perimental Area  near  timberline  in  the  Dischma  Valley  near 
Davos.  It  demonstrated  the  use  and  design  of  physical  barri- 
ers for  avalanche  control  and  reforestation  methods  with 
cembra  pine,  larch,  and  mugo  pine  for  long-term  reduction  of 
avalanche  problems. 

The  third  tour,  entitled  "Subalpine  and  Alpine  Environ- 
ments," emphasized  vegetation  succession,  particularly  of 
cembra  pine  and  larch  following  the  gradual  retreat  of  the 
Morteratsch  Glacier.  Also  included  in  this  tour  was  a  visit  to 
the  glacier  areas  at  Diavolezza  on  the  Bernina  Pass  where 
recreation  values  are  paramount.  From  the  aerial  tram  to  the 


mountain  station  (2,973  m)  we  viewed  the  strong  contrasts  of 
vegetation  and  glaciers  on  the  north  slopes  of  the  Bernina 
Pass  area. 

For  all  the  hosts  of  the  enlightening  field  tours  and  arrange- 
ments at  St.  fvloritz  we  owe  a  special  thanks.  Those  included 
Drs.  W.  Schonenberger,  R.  Hasler,  J.  Senn,  R.  Lassig, 
F.-K.  Holtmeier,  and  Herr  U.  Wasem — also  Engadine 
Foresters  J.  Altman,  I.  A.  Bisaz,  and  G.  Bott.  We  also  thank 
Prof.  Dr.  H.  Steinlin  from  the  Binding-Siftung  who  attended 
the  workshop  as  representative  of  the  Binding  Foundation. 
All  of  these  people  and  their  staffs  were  wonderful  hosts  who 
contributed  so  much  to  the  success  of  this  first  international 
conference  on  stone  pines. 

A  special  thanks  goes  to  Kathy  McDonald,  Intermountain 
Research  Station,  U.S.A.,  who  processed  all  the  manuscripts. 

The  organizing  committee  was  composed  of  the  following 
people: 

Dr.  W.  Schonenberger 
Dr.  R.  Hasler 
Dr.  J.  Senn 
Swiss  Federal  Institute  for 
Forest,  Snow  and  Landscape  Research 
Zurcher  Str.  1 1 1 
CH-8903  Birmensdorf 
Switzerland 

Prof.  Dr.  F.-K.  Holtmeier 

University  of  f^unster 
Department  of  Geography 
Landscape  Ecology  Division 
Robert-Koch-Str.  26 
W-4400  Munster 
Germany 

Dr.  Wyman  C.  Schmidt 
USDA  Forest  Service 
Intermountain  Research  Station 
Forestry  Sciences  Laboratory 
Montana  State  University 
Bozeman,  MT,  59717-0278,  U.S.A. 


Drs.  Holtmeier  and  Schmidt  served  as  technical  evaluators 
of  all  the  papers. 

As  a  consequence  of  this  international  workshop,  a  resolu- 
tion was  passed  by  the  scientists  in  attendance  that  aimed  to- 
ward greater  coordination  and  sharing  of  research  findings. 
The  resolution  was: 

1 .  To  organize  an  International  Association  for  Stone  Pine 
Research. 

2.  To  form  an  international  organizing  committee  by  elec- 
tion that  could  probably  meet  in  Tomsk,  Russia,  in  1993  (or 
later)  to  (a)  establish  the  objectives  of  the  Association;  and 
(b)  establish  the  organizational  structure  of  the  Association. 

3.  The  organizing  committee  to  develop  the  preliminary 
plan  for  the  next  International  Stone  Pine  Workshop. 

4.  To  hold  the  next  International  Stone  Pine  Workshop  at 
Tomsk,  Russia,  in  1995  (or  later). 

The  world  faces  many  environmental  and  resource  chal- 
lenges. Many  of  these  challenges  are  in  temperate  forests  of 
the  Northern  Hemisphere.  In  these,  stone  pines  are  a  signifi- 
cant forest  component  and  one  in  which  our  knowledge  base 
is  certainly  not  complete.  Sharing  the  knowledge  developed 
from  similar  ecosystems  and  building  on  that  solid  base  is 
key  to  a  better  understanding  of  how  our  subalpine  forests 
function  ecologically.  We  can  learn  much  from  each  other. 
The  international  workshop  that  generated  this  proceedings  is 
a  good  first  step  in  that  direction.  Let  us  do  what  we  can  to- 
gether to  keep  the  momentum. 


Wyman  C,  Schmidt  F.-Karl  Holtmeier 


VORWORT 


Bei  den  Arten  handelte  es  sich  um: 


Dieser  Band  ist  eine  erste  umfassende  internationale  Studie 
iiber  die  subalpinen  Steinkiefern  der  Erde.  Er  enthalt  die 
Vortrage,  die  anipiich  des  Internationalen  Workshops  "Subal- 
plne  Stone  Pines  and  Their  Environment — The  Status  of  Our 
Knowledge"  gehalten  wurden.  Dieser  Workshop  fand  vom 
5.  bis  1 1 .  Septennber  1 992  in  St.  Moritz  (Schweiz)  statt.  Die 
funf  subalpinen  Steinkieferarten  sind  ausschliepiich  in  der 
Nordhemisphare  verbreitet — davon  eine  in  Nordamerika 
und  vier  in  Eurasien.  Diese  Arten  weisen  viele  gemeinsame 
Merkmale  auf  und  besetzen  auch  uhnliche  okologische 
Nischen.  Unter  Steinkiefern  versteht  man  die  Arten  des 
Genus  Pinus  mit  funfnadeligen  Kurztrieben,  flugellosen 
Samen  und  Zapfen,  die  im  Reifezustand  geschlossen 
bleiben. 

Die  Vorgeschichte  dieses  Internationalen  Workshops 
beginnt  1 987.  Damals  fand  in  Bozeman  ein  Workshop  uber 
die  Whitebark  Pine  {Pinus  albicaulis)  statt,  dem  1989  ein 
gropes  Symposium,  ebenfalls  in  Bozeman,  folgte.  Beide 
Veranstaltungen  stiepen  auf  eine  grope  Resonanz  und  fuhrten 
zu  einem  wachsenden  Interesse  an  den  Hochgebirgsoko- 
systemen  des  westlichen  Nordamerika.  Eine  Vortragssektion 
dieses  Symposiums  war  den  Hochgebirgsokosystemen  der 
Welt  gewidmet.  Dabei  zeigte  sich,  dap  aus  einem  Vergleich 
mit  Steinkiefer-Okosystemen  anderer  in  ihrer  Naturaus- 
stattung  mit  dem  nordamerikanischen  Verbreitungsareal  der 
Whitebark  Pine  vergleichbarer  Regionen  der  Erde  auch 
wichtige  Informationen  zum  besseren  Verstandnis  der 
Whitebark  Pine  -  Walder  gewonnen  werden  konnen.  Als  ein 
wesentlicher  Unterschied  zu  den  amerikanischen  Whitebark 
Pine  -  Waldern  erwies  sich  die  zum  Tell  schon  einige  tausend 
Jahren  zuruckreichende  anthropogene  Beeinflussung  der 
eurasiatischen  Steinkiefernwalder  durch  den  Menschen.  Mit 
Sicherheit  lassen  sich  aus  den  dortigen  Verhaltnissen  auch 
nutzliche  Hinweise  auch  potentielle  Auswirkungen  men- 
schlicher  Eingriffe  in  den  amerikanischen  Whitebark  Pine  - 
Waldern  gewinnen. 

Anfang  1990  traf  sich  dann  eine  kleine  Gruppe  von  Wissen- 
schaftlern  und  Forstleuten  aus  den  westlichen  Vereinigten 
Staaten  mit  Dr.  Holtmeier  (Munster,  Deutschland)  in  Missoula 
(Montana),  um  uber  weiterfuhrende  Forschungen  iiber  die 
subalpinen  Steinkiefern  zu  beraten.  Der  nachste  Schritt 
sollte  eine  Verbesserung  und  Intensivierung  der  wissen- 
schaftlichen  Zusammenarbeit  der  Fachleute  aus  Europa, 
Asien  und  Nordamerika  sein.  Man  kam  uberein,  ein  en 
Internationalen  Workshop  zu  veranstalten,  mit  dem  Ziel,  dort 
Fragen  von  gemeinsamem  Interesse  zu  diskutieren  und  neue 
Forschungsansatze  zu  entwickeln.  Unmittelbar  nach  diesem 
Treffen  begannen  die  Vorbereitungen  zu  diesem  Workshop, 
der  dann  zweieinhalb  Jahre  spater  in  St.  Moritz  (Schweiz) 
stattfand. 

Seine  Hauptziele  waren: 

1 .  Verbesserung  und  Intensivierung  der  internationalen 
Zusammenarbeit. 

2.  Austausch  von  Forschungsergebnissen. 

3.  Forschungsbedarf  hinsichtlich  der  Okologie  und 
Bewirtschaftung  subalpiner  Steinkiefernwalder  aufzuzeigen. 

Uber  alle  funf  Steinkieferarten  wurden  Vortrage  gehalten. 
Insgesamt  waren  es  48  Referate,  an  denen  Wissenschaftler 
aus  12  Staaten  beteiligt  waren. 


Art 

Arve  (Synonym:  Zirbe) 
{Pinus  cembra) 

Sibirische  Arve 
{Pinus  sibirica) 


Sibirische  Zwergarve 
(Japanische 
Steinkiefer) 
{Pinus  pumila) 

Koreanische  Steinkiefer 
{Pinus  lioraiensis) 

Weiprinden-Kiefer 
{Pinus  albicaulis) 


Naturliches 
Verbreitunggebiet 

Europaische  Alpen,  Karpaten 


Russisch  Sibirien  und  ndrdliche 
Mongolei  (95  percent  in 
Rupiand;  5  percent  in  der 
Mongolei) 

Japan,  Korea,  Ostsibirien 


Nordost-China,  Nordkorea, 
Honshu,  Siidostasien 

Westliches  Nordamerika 


Folgende  Lander  waren  vertreten  (Zahl  der  Beitrage  in 
Klammern):  Rupiand  (9),  China  (1),  Japan  (4),  Schweiz  (5), 
Osterreich  (4),  Italien  (3),  Frankreich  (1),  Yugoslawien  (1), 
Tschechoslowakei  (1),  Rumanien  (1),  Deutschland  (3)  und 
USA  (15) 

Die  Hauptthemen  waren: 

•  Evolution  und  Taxonomie. 

•  Autokologie  der  subalpinen  Steinkieferarten. 

•  Wachstumscharakteristik. 

•  Einflup  der  Standortfaktoren. 

•  Regeneration. 

•  Bedeutung  der  Steinkiefern  fiir  die  Tiere. 

•  Waldstrukturen  und  -dynamik. 

•  Waldbewirtschaftung  und  -pflege. 

•  Forschungsbedarf. 

Die  Steinkiefern  stellen  einen  hohen  Anteil  an  der  Wald- 
bedeckung  groper  Gebiete.  Allein  Pinus  sibirica  bedeckt  ein 
Areal  von  40  Millionen  Hektar.  Der  gropte  Teil  davon  liegt  in 
Sibirien,  wo  der  Sibirischen  Arve  sowohl  in  ihrer  Bedeutung 
fur  die  Wirtschaft  als  auch  im  Hinblick  auf  die  biologische 
Situation  die  Rolle  eines  Schlusselfaktors  zukommt.  Pinus 
koraiensis  stellt  ein  Viertel  der  Holzproduktion  in  China. 
Pinus  albicaulis  ist  unter  anderem  fur  den  Grizzly-Bar  von 
existentieller  Bedeutung.  Auperdem  erfullt  diese  Art  in  hoch- 
gelegenen  Wassereinzugsgebieten  eine  wichtige  Schutz- 
funktion  und  erhoht  zudem  wesentlich  den  asthetischen  Reiz 
der  Gebirgslandschaft.  Pinus  cembra  kommt  in  Europa  eine 
wichtige  Rolle  bei  der  Lawinenvorbeugung  durch  Schutz- 
walder  zu.  Zugleich  liefert  sie  ein  fiir  Tafelungen  und  Schnitz- 
erein  geschatztes  Holz.  Auch  sie  ist  von  hohem  asthetischen 
Wert.  Pinus  pumila  tragt  in  Japan  und  Kamtschatka  ent- 
scheidend  zur  Stabilisierung  steiler  Gebirgshange  bei.  Die 
vielen  Gemeinsamkeiten  der  genannten  Arten  haben  ihre 
Ursache  vermutlich  in  ihrer  Abstammung  von  einer  Art, 
moglichenweise  von  Pinus  sibirica.  Charakteristisch  ist  fiir  sie 
auch  die  Verbreitung  ihrer  Samen  durch  den  nordamerika- 
nischen und  europaischen  Tannenhaher.  Zwischen  den 
subalpinen  Steinkiefern  und  den  Hahern  hat  sich  im  Laufe 
der  Evolution  ein  echter  Mutualismus  entwickelt. 


Der  Workshop  wurde  unterstutzt  durch  die  Intermountain 
Research  Station,  USDA  Forest  Service,  USA,  die  West- 
falische  Wilhelms-Universitat  (MOnster,  Deutschland),  die 
Eidgenossische  Forschungsanstait  fur  Wald,  Schnee  und 
Landschaft  (Schweiz)  sowie  durch  die  Karl  und  Sophie  Bind- 
ing -  Stiftung. 

Wir  danken  alien  Personen  und  Organisationen,  die  mit 
ihrem  Enthusiasmus  und  ihrer  tatkraftigen  Unterutzung  die 
Planung  und  die  Durchfuhrung  dieses  Workshops  ermbglicht 
haben. 

Wahrend  des  Workshops  fanden  unter  der  Leitung  der 
Gastgeber  drei  Exkursionen  statt.  Die  erste  fuhrte  in  die 
Larchen-Arvenwalder  oberhalb  des  bekannten  Engadiner 
Hohenkurortes  St.  Moritz.  Hauptthemen  waren  die  Ver- 
jungung  und  Sukzession  in  Lawinenzugen  und  im  Wald- 
grenzbereich,  die  Einbindung  von  Eichhornchen  und  Tannen- 
haher  in  das  Okosystem  des  Larchen-Arvenwaldes,  die 
Einflusse  von  Beweidung  und  Wild  sowie  die  Belastungen 
der  Walder  und  der  Landschaft  durch  den  modernen 
Tourismus. 

Die  zweite  Exkursion,  die  sich  im  wesentlichen  mit  okologi- 
schen  Problemen  und  Techniken  bei  der  Hochlagenauf- 
forstung  und  der  Behandlung  von  Lawineneinzugsgebieten 
befapte,  fuhrte  in  die  im  Waldgrenzbereich  gelegenen  Ver- 
suchsflachen  am  Stillberg  im  Dischmatal  bei  Davos.  Dort 
wurden  kunstliche  Lawinenverbauungen  und  Wieder- 
aufforstungen  mit  Arven,  Larche  und  Legfohre  als  Mittel  zur 
Lawinenvorbeugung  demonstriert. 

Die  dritte  Tour,  die  unter  dem  Thema  "subalpine  und  alpine 
Lebensraume"  stand,  fuhrte  zunachst  in  das  Morteratsch-Tal. 
Dort  ging  es  vor  allem  um  die  Vegetationssukzession  im 
Vorfeld  des  allmahlich  zuruckweichenden  Morteratsch- 
Gletschers  unter  besonderer  Berucksichtigung  der 
Besiedlung  durch  Arve  und  Larche.  Mit  einem  Besuch  der 
Diavolezza  am  Berninapap  fand  die  Exkursion  ihren 
Abschlup.  Von  der  Bergstation  (2973  m)  der  Luftseilbahn  bot 
sich  ein  uberwaltigender  Blick  auf  die  stark  vergletscherte 
Nordseite  der  Berninagruppe. 

Fur  alles,  die  eriebnisreichen  und  lehrrreichen  Exkur- 
sionen sowie  die  Veranstaltungen  in  St.  Moritz,  schulden  wir 
den  Gastgebern  —  Drs.  W.  Schonenberger,  R.  Hasler, 
J.  Senn,  R.  Lassig,  F.-K.  Holtmeier  und  Herr  U.  Wasem  — 
sowie  den  Engadiner  Fbrstern  J.  Altmann,  I. A.  Bisaz,  und 
G.  Bott  besonderen  Dank.  Nicht  zuletzt  geht  unser  Dank 
auch  an  Herrn  Prof.  Dr.  H.  Steinlin,  der  als  Vertreter  der 
Binding-Stiftung  selbst  an  dem  Workshop  teilnahm.  Sie  alle 
haben  mit  ihrem  gro^en  Engagement  bei  der  Planung  und 
der  Durchfuhrung  zum  Gelingen  dieses  ersten  internationalen 
Workshops  uber  die  Steinkiefern  beigetragen. 

Besonderer  Dank  gilt  auch  Frau  Kathy  McDonald  und 
Herrn  Richard  Klade  (Intermountain  Research  Station,  USA) 
fur  die  redaktionelle  Bearbeitung  der  Manuskripte. 

Das  Organistaionskomitee  setzte  sich  zusammen  aus: 

Dr.  W.  Schonenberger 
Dr.  W.  Hasler 
Dr.  J.  Senn 
Eidgenossische  Forschungsanstait 
fur  Wald,  Schnee  und  Landschaft 
Zurcher  Strape  1 1 1 
Ch-8093  Birmensdorf 
Schweiz 


Prof.  Dr.  F.-K.  Holtmeier 
Westfalische  Wilhelms-Universitat 
Institut  fur  Geographie 
Abteilung  Landschaftsokologie 
Robert-Koch-Strape  29 
D-48149  Munster 
Deutschland 

Dr.  Wyman  C.  Schmidt 
USDA  Forest  Service 
Intermountain  Research  Station 
Forestry  Sciences  Laboratory 
Montana  State  University 
Bozeman,  Montana,  USA  59717-0278 

Die  Vorbereitung  der  Manuskripte  fur  die  redaktionelle 
Bearbeitung  wurde  von  Dr.  Holtmeier  und  Dr.  Schmidt 
durchgefuhrt. 

Wahrend  des  Workshops  veranschiedeten  die  Teilnehmer 
eine  Resolution,  die  eine  noch  weitergehende  Koordination 
der  Forschung  und  bessere  gegenseitige  Information  uber 
neue  Forschungsergebnisse  zum  Ziel  hat.  Sie  umfapt 
folgende  Punkte: 

1 .  Grundung  einer  Internationalen  Vereinigung  fur  die 
Erforschung  der  Steinkiefern. 

2.  Bildung  eines  Internationalen  Organisationskomitees, 
das  erstmals  1993  (oder  spater)  in  Tomsk  (Rupiand) 
zusammentreten  soli,  um  die  kunftigen  Ziele  dieser  Ver- 
einigung zu  formulieren  und  eine  Organisationstruktur  zu 
schaffen. 

3.  Planung  des  nachsten  Internationalen  Workshops  uber 
die  Steinkiefern. 

4.  Durchfuhrung  dieses  Workshops  voraussichtlich  1995 
(Oder  spater)  in  Tomsk. 

Die  Welt  sieht  sich  vielen  Herausforderungen  durch 
Umweltprobleme  und  Ressourcennutzung  gegenuber. 
Davon  sind  auch  die  Gebirgswalder  und  borealen  Walder  der 
Nordhemisphare  betroffen,  zu  denen  auch  die  Steinkiefern- 
walder  gehoren.  Das  Wissen  uber  sie  ist  noch  luckenhaft. 
Indem  wir  unser  Wissen  zusammentragen  bilden  wir  eine 
solide  Grundlage,  von  der  aus  es  leichter  sein  wird,  tiefere 
Einsichten  in  die  okologischen  Wirkungsgefuge  dieser 
Waldokosysteme  zu  gewinnen.  Der  Internationale  Workshop 
in  St.  Moritz  mit  den  hier  zusammengestellten  Beitragen  ist 
ein  erster  Schritt  in  diese  Richtung.  Jetzt  geht  es  darum, 
diesen  Weg  weiterzugehen. 


Wyman  C.  Schmidt  F.-Kari  Holtmeier 


CONTENTS 


AUTHORS  TITLE  PAGE 

Wyman  C.  Schmidt  Distribution  of  Stone  Pines  1 

FIELD  TOURS 

Swiss  Hosts  Tour  Descriptions  8 

Mihailo  Grbic 

Friedrich-Karl  Holtmeier  Introduction  to  the  Upper  Engadine  and  Its  Forest  9 

EVOLUTION  AND  TAXONOMY 
Konstantin  V.  Krutovskli  Genetic  Differentiation  and  Phylogeny  of  Stone  Pine  Species  Based  on 

Dmitri  V.  Politov  Isozyme  Loci  19 

Yuri  P.  Altukhov 

Hermann  Mattes  Coevolutional  Aspects  of  Stone  Pines  and  Nutcrackers  31 

Dmitri  V.  Politov  Allozyme  Polymorphism,  Heterozygosity,  and  Mating  System  of  Stone  Pines  36 

Konstantin  V.  Krutovskii 

Diana  F.  Tomback  Genetic  Population  Structure  and  Growth  Form  Distribution  in  Bird-Dispersed  Pines  43 

William  S.  F.  Schuster 

BASIC  ECOLOGY 

Ernst  Frehner  Experiences  With  Reproduction  of  Cembra  Pine  52 

Walter  Schonenberger 

P.  L.  Gorchakovsky  Distribution  and  Ecology  of  Siberian  Stone  Pine  in  the  Urals  56 

Rudolf  Hasler  Ecophysiological  Investigations  on  Cembran  Pine  at  Timberline  in  the  Alps,  An  Overview  61 

Peter  A.  Khomentovsky  A  Pattern  of  Pinus  pumila  Seed  Production  Ecology  in  the  Mountains  of  Central 

Kamtchatka  67 

Takayoshi  Koike  Needle  Longevity  and  Photosynthetic  Performance  in  Cembran  Pine  and  Norway  Spruce 

Rudolf  Hasler  Growing  on  the  North-  and  East-Facing  Slopes  at  the  Timberline  of  Stillberg  in  the 

Hans  Item  Swiss  Alps  78 

Luo  Ju  Chun  The  Broad-leaved  Korean  Pine  Forest  in  China  81 

Tad  Weaver  Climates  Where  Stone  Pines  Grow,  A  Comparison  85 

GROWTH  CHARACTERISTICS 

Kurt  Holzer  Growth  of  Swiss  Stone  Pines  That  Originated  From  and  Were  Planted  at  Several 

Altitudes  in  the  Austrian  Alps  91 

Takuya  Kajimoto  Seasonal  Patterns  of  Growth  and  Photosynthetic  Activity  of  Pinus  pumila  Growing  on  the 

Kiso  Mountain  Range,  Central  Japan  93 

Herbert  Kronfuss  Height  Growth  in  Cembran  Pine  as  a  Factor  of  Air  Temperature  99 

Josef  Senn  Survival  and  Growth  of  Planted  Cembran  Pines  at  the  Alpine  Timberline  105 

Walter  Schbnenberg 
Ueli  Wasem 


AUTHORS  TITLE  PAGE 

INFLUENCE  OF  ENVIRONMENTAL  FACTORS 

Mihailo  Grbic  Changes  of  Swiss  Stone  Pine  Aphid  Life  Cycle,  Density,  and  Population  Structure  in 

High-Altitude  Swiss  Stone  Pine  Afforestation   111 

Raymond  J.  Hoff  Genetic  Consequences  and  Research  Challenges  of  Blister  Rust  in  Whitebark  Pine 

Susan  K.  Hagle  Forests  118 

Richard  G.  Krebill 

Karel  Kahik  Performance  of  Pinus  cembra,  P.  peace,  and  P.  strobiformis  Within  Air-Polluted  Areas   127 

Todd  Kipfer  Competition  and  Crown  Characteristics  of  Whitebark  Pine  Following  Logging  in 

Katherine  Hansen  Montana,  U.S.A   130 

Ward  McCaughey 

Penelope  Morgan  Fire  Ecology  of  Whitebark  Pine  Forests  of  the  Northern  Rocky  Mountains,  U.S.A   136 

Stephen  C.  Bunting 
Robert  E.  Keane 
Stephen  F.  Arno 

Tad  Weaver  Vegetation  Distribution  and  Production  in  Rocky  Mountain  Climates — With  Emphasis 

on  Whitebark  Pine  142 

REGENERATION 

I.  Blada  Variation  in  Size  and  Weight  of  Cones  and  Seeds  in  Four  Natural  Populations  of 

N.  Popescu  Carpathian  Stone  Pine  154 

Mitsuhiro  Hayashida  Role  of  Nutcrackers  on  Seed  Dispersal  and  Establishment  of  Pinus  pumila  and 

P.  pentaphylla  1 59 

Harry  E.  Hutchins  Role  of  Various  Animals  in  Dispersal  and  Establishment  of  Whitebark  Pine  in  the 

Rocky  Mountains,  U.S.A  163 

Hermann  Mattes  Size  of  Pine  Areas  in  Relation  to  Seed  Dispersal  172 

Ward  W.  McCaughey  The  Regeneration  Process  of  Whitebark  Pine   179 

Ikuko  Nakashinden  Japanese  Stone  Pine  Cone  Production  Estimated  From  Cone  Scars,  Mount 

Kisokomagatake,  Central  Japanese  Alps   188 

Diana  F.  Tomback  Effects  of  Seed  Dispersal  by  Clark's  Nutcracker  on  Early  Postfire  Regeneration  of 

Whitebark  Pine   193 

Vladislav  N.  Vorobjev  New  Trend  in  Dendrochronology:  1.  Theoretical  Principles  of  Reprochronology   199 

V.  N.  Vorobjev  New  Trend  in  Dendrochronology:  2.  Method  of  Retrospective  Study  of  Seminiference 

S.  N.  Goroshkevich  Dynamics  in  Pinaceae  201 

D.  A.  Savchuk 

IMPORTANCE  TO  WILDLIFE 

Ronald  M.  Lanner  Nutritive  Value  of  Whitebark  Pine  Seeds,  and  the  Question  of  Their  Variable  Dormancy  206 

Barrie  K.  Gilbert 

David  J.  Mattson  Bear  Use  of  Whitebark  Pine  Seeds  in  North  America  212 

Daniel  P.  Reinhart 

Diana  F.  Tomback  Ecological  Relationship  Between  Clark's  Nutcracker  and  Four  Wingless-Seed  Strobus 

Pines  of  Western  North  America  221 


AUTHORS  TITLE  PAGE 

FOREST  STRUCTURE  AND  DYNAMICS 

Giovanna  De  Mas  Structure  of  Swiss  Stone  Pine  Stands  in  Northeastern  Italy  226 

Elena  Piutti 

S.  Yu.  Grishin  Role  of  Pinus  pumila  in  Primary  Succession  on  the  Lava  Flows  of  Volcanoes  of 

Kamchatka  240 

Robert  E.  Keane  Decline  of  Whitebark  Pine  in  the  Bob  Marshall  Wilderness  Complex  of  Montana,  U.S.A  245 

Penelope  Morgan 

Renzo  Motta  Some  Aspects  of  Cembran  Pine  Regeneration  in  the  Italian  Cottian  Alps  254 

Alberto  Dotta 

Pietro  Piussi  Mixed  Cembran  Pine  Stands  on  the  Southern  Slope  of  the  Eastern  Alps  261 

Siegfried  Sauermoser  Current  Distribution  of  Cembra  Pine  in  the  Lechtal  Alps  269 

E.  P.  Smolonogov  Geographical  Differentiation  and  Dynamics  of  Siberian  Stone  Pine  Forests  in  Eurasia  275 

Use  WI6rick  The  Cembran  Pine  in  the  French  Alps:  Stand  Dynamics  of  a  Cembran  Pine  Forest 

in  Tueda  (Savoy,  France)  280 

FOREST  MANAGEMENT 

Werner  Frey  Silvicultural  Treatment  and  Avalanche  Protection  of  Swiss  Stone  Pine  Forests  290 

Riet  Gordon  Importance  and  Silvicultural  Treatment  of  Stone  Pine  in  the  Upper  Engadine  (Grisons)  294 

Jorg  Heumader  Cultivation  of  Cembran  Pine  Plants  for  High-Elevation  Afforestations  298 

Katherine  C.  Kendall  Whitebark  Pine  Conservation  in  North  American  National  Parks  302 

RESEARCH  NEEDS 

Wyman  C.  Schmidt  Research  Needs  in  Whitebark  Pine  Ecosystems  309 

Vladislav  N.Vorobjev  Problems  of  Comprehensive  Investigation,  Utilization,  and  Reproduction  of 

Nina  A.  Vorobjeva  Russian  Cedar  Pine  Forests  314 


DISTRIBUTION  OF  STONE  PINES 

Wyman  C.  Schmidt 


The  five  stone  pine  species  that  are  the  subject  of  this 
proceedings  occur  only  in  the  Northern  Hemisphere — one 
species  in  North  America  and  the  other  four  in  Europe 
and  Asia.  Taxonomically  they  fall  in  section  Strobus  and 
subsection  Cembrae.  They  are  characterized  by  having 
five-needled  fascicles,  cones  that  do  not  open  at  maturity, 
and  seeds  that  are  wingless. 

The  distributions  of  the  five  species  are  reasonably  well 
defined,  but  updates  on  their  distribution  usually  show 
minor  changes.  Also,  in  some  cases,  such  as  with  white- 
bark  pine  (Pinus  albicaulis),  the  species  occurs  only  at 
high-elevation  sites,  and  there  may  be  long  distances  be- 
tween the  mountain  ranges  that  support  this  species.  This 
species  ocoirs  over  a  large  geographic  area;  however,  there 
are  usually  large  intervening  valleys  and  lowlands  not 
suitable  for  whitebark  pine  forests.  The  same  is  true  for 
the  other  stone  pines.  Therefore,  just  because  a  distribu- 
tion boundary  of  a  species  encompasses  a  large  area,  it 
does  not  mean  that  all  the  mountains  and  valleys  are 
fully  occupied  by  that  species. 

For  the  purposes  of  this  proceedings,  we  have  chosen  to 
show  the  general  distribution  on  the  maps  of  the  five  spe- 
cies and  do  not  necessarily  include  all  of  the  individual 
outliers. 

The  distribution  maps  that  follow  are  adapted  fi'om  those 
pubUshed  by  Critchfield  and  Little  (1966).  The  maps  show 
that  the  five  species  are  mostly  discrete  populations,  but 
there  are  overlapping  distributions  of  Siberian  stone  pine 
(Pinus  sibirica)  and  Japanese  stone  pine  (P.  pumila),  and 
an  overlap  of  Japanese  stone  pine  and  Korean  stone  pine 
(P.  koraiensis). 

Whitebark  pine  occurs  only  in  North  America  at  high 
elevations  fi*om  central  British  Columbia  south  through 
the  Cascade  Mountains  of  Washington  and  Oregon  to  the 
southern  Sierra  Nevada  of  California,  in  a  limited  area  of 
V   Nevada,  and  extensively  in  Montana,  Idaho,  and  western 
Wyoming.  It  occurs  just  below  the  timberline  and  extends 
downward  elevationally  into  associations  with  several 
other  conifers. 

Swiss  stone  pine  grows  at  high  elevations  in  the  Alps  and 
Carpathian  Moimtains.  It  occurs  mostly  in  Switzerland, 
Austria,  northern  Italy,  southeastern  France,  Romania, 
Czechoslovakia,  and  Poland.  Although  Swiss  stone  pine 
is  considered  closely  related  to  Siberian  stone  pine,  over 
2,000  kilometers  presently  separate  the  two  species. 


Swiss  stone  pine  commonly  grows  in  association  with 
other  conifers. 

Siberian  stone  pine  is  widely  distributed  in  Eurasia  fi"om 
the  Ural  Moimtains  through  western  and  central  Siberia  to 
northern  MongoUa — an  east-west  distance  of  about  4,000 
kilometers.  It  overlaps  the  range  of  Japanese  stone  pine 
on  the  eastern  portion  of  its  distribution  near  Lake  Baikal 
in  Siberia.  Different  than  both  whitebark  pine  and  Swiss 
stone  pine,  which  only  occupy  high-elevation  sites,  Sibe- 
rian stone  pine  (usually  called  cedar-pine  or  Siberian  ce- 
dar by  Russians)  commonly  grows  in  the  plains  and  river 
valleys,  as  well  as  in  the  mountain  areas. 

Japanese  stone  pine  grows  in  northeastern  Siberia  and 
in  subalpine  parts  of  Japan,  Korea,  and  Manchvuia.  It  ex- 
tends north  almost  to  the  Arctic  Ocean  (70°  N.),  west  to 
Mongolia  and  to  the  Lake  Baikal  area,  and  south  to  Korea 
and  Honshu,  Japan.  It  grows  from  low  elevations  in  the 
north  to  high  elevations  in  the  south.  It  commonly  forms 
thickets  just  below  barren  tundra  on  tops  of  mountains  or 
lowlands  in  the  extreme  north.  Like  Siberian  stone  pine, 
it  extends  nearly  4,000  kilometers  fi*om  east  to  west  and 
3,000  kilometers  north  to  south. 

Korean  stone  pine  occurs  throughout  Korea  and  eastern 
Manchuria  into  southeastern  Siberia  as  well  as  on  the  is- 
lands of  Honshu  and  Shikoku  in  Japan.  It  tends  to  grow 
in  more  maritime  conditions  and  lower  moimtains  than  its 
neighbor — the  Japanese  stone  pine.  Its  geographic  range 
overlaps  that  of  the  Japanese  stone  pine,  but  the  species 
occupy  different  ecological  niches.  Korean  stone  pine,  im- 
like  its  counterparts,  will  grow  in  association  with  hard- 
woods where  near-maritime  cUmates  prevail.  Its  valuable 
wood  makes  it  a  much  sought  after  species. 

The  following  maps  illustrate  the  geographic  distribu- 
tion of  the  five  stone  pines.  These  descriptions  are  based 
primarily  on  those  of  Critchfield  and  Little  (1966)  and 
Mirov  (1967). 

References 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agricvdture,  For- 
est Service.  97  p. 

Mirov,  N.  T.  1967.  The  genus  Pinus.  New  York:  Ronald 
Press  Company.  602  p. 


1 


NORTH  AMERICA 


135°  130°  125°  120°  115°  110° 


130°  125°  120°  115°  110°  105°  100° 


2 


EUROPE 


10°  5°  0°  5°  10°        15°  20°  25°  30°  35° 


0°  5°  10°  15°  20°  25°  30°  35° 


3 


ASIA 


4 


ASIA 


5 


ASIA 


Field  Tours 


International  Workshop 
St.  Moritz  1 992 


Field  tours  were  an  important  component  of  the  Stone  Pine  Workshop.  First, 
the  tours  provided  an  informal  setting  for  scientists  from  the  12  countries  to  become 
acquainted  with  each  other  and  their  research;  second,  the  tours  demonstrated  the 
research  and  management  of  Swiss  stone  pine  forests  in  the  rugged  and  spectacularly 
beautiful  mountains  of  Switzerland. 

The  following  section  is  the  handout  from  the  Swiss  hosts  used  to  describe 
the  three  major  field  tours.  The  drawings  by  a  gifted  scientist,  Mihailo  Grbic,  from 
Belgrade,  Yugoslavia,  depict  a  symbolic  participant,  Dr.  Stone  Pine,  enjoying  the  rather 
strenuous  excursions.  Grbic  also  created  the  welcome  drawing  used  as  sections 
division  indicators  in  this  publication. 


7 


TOUR  DESCRIPTIONS 

Swiss  Hosts 


We  will  try  to  use  pleasant  weather  for  the  field  tours. 
Therefore  this  preliminary  schedule  may  change  a  little. 
Usually  weather  is  fine  in  September.  However,  snow  and 
fi-ost  are  possible.  Do  not  forget  warm  clothing,  tough  foot- 
wear, backpack,  sunblock,  and  simglasses.  During  the  field 
toiu-s  we  will  have  lunch  fi-om  the  backpack. 

Engadine  Cembra  Pine-Larch  Forest 

Sunday  6th.  St.  Moritz,  Muottas  da  Schlarigna.  Walk 
fi-om  St.  Moritz  Bad  (Hotel  "Laudinella'')-Stazer  Wald- 
Muottas  da  Schlarigna-St.  Moritz  Bad  (1,800  to  2,300  m 
above  sea  level  [a.s.l.]). 

Cembra  pine-larch  forest:  composition,  structure,  growth 
characteristics,  old  cembra  pines,  regeneration  pattern, 
understory  vegetation,  avalanche  tracks  with  green  alder 
{Alnus  viridis)  formation,  influence  of  seed-caching  ani- 
mals (squirrel,  nutcracker),  effects  of  seed  dispersal  by 
the  thick-billed  European  nutcracker  (Nucifraga  c. 
caryocatactes). 

Human  influence  on  the  forest  distribution  pattern, 
human-caused  timberline,  influence  of  grazing  livestock 
and  game,  and  of  toiu"ism;  invasion  of  rarely  used  and 
abandoned  alpine  and  forest  pastures  by  trees,  influence 
of  microtopography  and  microclimate  on  alpine  vegetation 
and  on  growth  of  invading  trees. 

Ecology  and  Technique  of  High- 
Altitude  Afforestation 

Tuesday  8th.  Davos,  Stillberg.  Bus  fi"om  St.  Moritz- 
Albula  pass-Davos-Teufi  (Dischma  Valley,  1,700  m), 
5-hour  walk  to  and  through  Stillberg  experimental  area 
near  timberline  at  2,000  to  2,230  m  a.s.l.  and  back  to  the 
valley  bottom  (Teufi). 

The  experimental  area  lies  on  a  steep  avalanche  catch- 
ment slope.  Construction  of  avalanche  barriers.  Dam- 
ages, diseases,  mortality  and  development  of  a  17-year-old 


plantation  of  Pinus  cembra,  Larix  decidua,  and  P.  mugo. 
Influence  of  different  site  factors  such  as  wind,  tempera- 
ture, irradiation,  snow  movement,  duration  of  snow  cover, 
soil  properties,  vegetation  on  plantation  success.  Eco- 
physiology  of  subalpine  timberline  trees.  Conclusions  for 
plantation  technique  and  spatial  patterns. 


Subalpine  and  Alpine  Environments 

Friday  11th.  Morteratsch  Glacier,  Diavolezza. 
Morteratsch  Glacier  (2,000  m,  2-  to  3-hour  walk).  Train  to 
Morteratsch  station,  walk  to  Morteratsch  Glacier  via  Alp 
Chiinetta  (2,052  m);  old-growth  cembra  pine  forests,  re- 
generation, invasion  of  the  lateral  moraine  (fi-om  1850) 
and  the  glacier  forefield  by  cembra  pine  and  larch. 

Diavolezza  (2,973  m).  Train  fi-om  Morteratsch  station 
through  the  Bemina  Valley  to  Diavolezza  station  (2,200  m) 
on  Bemina  Pass,  and  by  ropeway  to  Diavolezza.  Magnifi- 
cent view  of  the  Bernina  Massive  (4,049  m)  with  Pers  Gla- 
cier and  Morteratsch  Glacier. 

★  ***★ 

In  addition  to  providing  valuable  scientific  and  manage- 
ment information,  the  introductory  paper  by  Friedrich- 
Karl  Holtmeier  that  follows  gave  background  and  baseline 
data  describing  the  local  Swiss  stone  pine  forests  that 
were  the  focus  of  the  field  tours. 

***** 


8 


INTRODUCTION  TO  THE  UPPER 
ENGADINE  AND  ITS  FOREST 

Friedrich-Karl  Holtmeier 


Abstract — The  Engadine  is  characterised  by  great  mean  eleva- 
tion (2,400  m)  and  a  relatively  continental  climate.  Cembra  pine 
{Pinus  cembra)  and  European  larch  (Larix  decidua)  are  the  most 
important  tree  species.  The  area  was  already  settled  in  prehis- 
toric time.  Since  then,  mountain  forest  has  been  influenced  by 
human  impact.  During  history  timberline  became  lower  by  about 
150-300  m.  At  present,  the  upper  forest  limit  is  located  at  about 
2,200-2,300  m.  Composition  and  structure  of  the  forest  also 
changed  in  that  larch  could  spread  at  the  cost  of  cembra  pine  on 
all  areas  that  were  suitable  for  grazing.  After  grazing  pressure 
has  declined  natural  succession  is  revived,  and,  in  the  long  term, 
larch  will  be  replaced  by  cembra  pine,  if  not  specially  managed. 
Natural  stands  of  Norway  spruce  (Picea  ahies)  occur  on  the 
northwest-exposed  slopes  of  the  main  valleys  within  an  area 
characterized  by  higher  humidity.  Spruce  reaches  its  upper  limit 
at  2,000  m.  Dwarfed  mountain  pine  (Pinus  mugo)  and  green  al- 
der (Alnus  viridis)  are  most  common  in  avalanche  tracks.  Green 
alder  is  to  be  found  mainly  on  moist  sites,  whereas  mountain 
pine  prevails  under  dry  conditions.  Alder  also  occurs  in  bogs. 


The  Engadine — Engiadina  in  Rhacto-Romsinic  lan- 
guages— is  located  in  the  eastern  part  of  Switzerland  and 
comprises  the  uppermost  drainage  area  of  the  Inn  River 
(fig.  1).  The  main  valley  trends  from  southwest  to  north- 
east. By  an  administrative  border  the  Engadine  is  divided 
into  Upper  and  Lower  Engadine.  However,  landscape 
physiognomy  of  both  sections  is  also  different.  The  Lower 
Engadine  is  deeply  cut  into  the  mountains,  and  mountain 
slopes  are  steep  and  nagged.  The  Upper  Engadine  main 
valley,  on  the  other  hand,  is  wide  and  topography  is  com- 
paratively smooth.  Between  St.  Moritz  and  the  Maloja 
Pass  the  valley  is  characterized  by  four  lakes  separated  by 
alluvial  plains.  To  the  southeast  the  Bemina  Valley  gives 
access  to  the  famous  glaciated  Bemina  Group. 

The  Engadine  is  characterized  by  great  mean  elevation 
(2,400  m).  The  bottom  of  the  main  valley  is  located  at  an 
altitude  of  about  1,800  m.  The  tributary  valleys  climb  up 
to  about  2,300  m.  The  highest  peaks'  elevations  are  about 
4,000  m  (Piz  Bernina  4,049  m;  fig.  1).  In  view  of  its  great 
elevation  the  Upper  Engadine  has  been  called  The  Roof 
of  Europe"  by  the  English. 

Due  to  the  geographical  location  in  the  central  Alps 
and  the  high  mass-elevation  of  the  Engadine  the  climate 
is  rather  continental,  characterized  by  a  relatively  great 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Friedrich-Karl  Holtmeier  is  Professor  of  Landscape  Ecology  and  Geogra- 
phy, Department  of  Geography,  University  of  Munster,  Germany. 


amplitude  of  mean  temperature  and  comparatively  low 
amount  of  precipitation  (fig.  2).  Annual  precipitation  de- 
creases from  about  1,400  mm  at  Maloja  Pass,  where  the 
main  valley  is  open  to  moisture-carrying  air  ciurents  from 
the  southwest  (fig.  3),  to  less  than  1,000  mm  at  St.  Moritz. 
At  timberline,  however,  precipitation  is  much  higher  than 
at  the  valley  bottom,  as  can  be  concluded  from  the  data 
available  for  Julier  Hospiz  on  Julier  Pass  and  for  Bemina 
Hospiz  on  Bemina  Pass.  At  both  stations,  anniial  precipi- 
tation exceeds  1,500  mm. 

Due  to  the  continental  climate,  the  upper  limits  of  veg- 
etation and  cultivation  reach  relatively  high  altitudes.  In 
the  Upper  Engadine,  forest  limit  is  located  at  about  2,200- 
2,300  m.  Solitary  crippled  trees  may  still  be  found  at  and 
even  above  2,500  m  (Holtmeier  1965,  1974). 

In  the  Upper  Engadine,  artifacts  were  found  that  give 
evidence  of  agriculttire  during  the  Bronze  Age  (about 
3,000  years  B.P.),  while  in  the  Lower  Engadine  hunters 
and  shepherds  were  already  present  in  the  Neohthic  Age 
(4,000  years  ago;  Conrad  1940).  Trading  and  mining  be- 
came increasingly  important  to  the  economy  dvuing  his- 
toric time.  Thus,  there  was  increasing  human  impact  on 
the  moimtain  ecosystems  throughout  history.  Almost  no 
untouched  nature  was  left,  although  the  kind  and  the  in- 
tensity of  the  himaan  disturbances  were  different. 

DEPRESSION  OF  TIMBERLINE 

Due  to  the  great  elevation  of  the  area  humans  could 
only  use  the  mountains  up  to  a  certain  altitude;  there  the 


Figure  1— Location  of  the  Upper  Engadine. 


9 


3000 


2500 


2000 


1500 


1000  - 


500  - 


mm 


Figure  2 — Annual  precipitation.  Note  the  de- 
crease from  Maloja  Pass  toward  St..  Moritz  and 
Schuls  (Lower  Engadine). 


Figure  4 — Zuoz  Village  (1 ,700  m).  The  original 
forest  was  almost  totally  removed  as  a  conse- 
quence of  alpine  pasturing  (upper  slope)  and 
agriculture  (artificial  terraces  above  the  village). 
Only  two  small  stands  were  left  on  the  steep 
north-exposed  slopes  of  the  narrow  tributary 
valleys  and  one  above  the  village  for  protection 
from  avalanches.  Photo  taken  1 968. 


main  influence  was  concentrated  in  the  forest  belt  not 
only  on  favorable  southern  exposures  but  also  on  all 
sections  that  were  fairly  accessible — trough  shoulders, 
terraces,  and  similar  almost-level  areas  even  in  north- 
exposed  locations.  During  former  times  cereals  were 
grown  up  to  elevations  of  about  2,300  m,  favorable  topog- 
raphy provided.  Many  of  the  high-altitude  forests  were 
cleared  for  pastoral  use,  while  those  at  lower  elevations 


gave  way  to  agriculture.  The  extent  of  timberline  depres- 
sion varied  in  relation  to  exposure  and  topography.  On 
some  easily  accessible  southern  exposures  of  the  main 
valley,  for  example,  and  of  the  Fex  Valley  and  Fedoz  Val- 
ley, the  forest  was  totally  removed  except  for  sections 
characterized  by  too  great  steepness  (figs.  4,  5).  Other  for- 
ests were  destroyed  by  ore  mining,  salt  works,  and  char- 
coal production,  especially  during  the  Middle  Ages.  On 


Figure  3 — View  from  Lej  dais  Chods  (northwest- 
facing  slope  of  the  main  valley)  southwest  to- 
ward the  gap  of  Maloja  Pass. 


Figure  5 — View  to  southeast-exposed  slope  of  the  main 
valley.  In  the  foreground,  Samaden  Village  (1 ,720  m). 
Forest  was  removed  from  all  gentle  topography  (trough 
shoulders,  lower  slope  area)  and  reduced  to  a  narrow 
belt  on  the  steep  trough  wall.  Photo  taken  1969. 


10 


Figure  6 — View  from  Diavolezza  trail  to 
Bernina  Pass  (2,328  m).  Bernina  Pass  was 
covered  by  larch-cembra  pine  forest  until  the 
Middle  Ages.  Then,  the  forest  became  a 
victim  of  mining.  Until  present,  the  area  is 
used  in  summer  for  grazing  cattle. 


the  Bernina,  Abula,  and  Jiilier  Passes  (about  2,300  m)  the 
forest  disappeared  as  a  consequence  of  mining.  These  ar- 
eas v^^ere  then  used  as  alpine  pastures  (fig.  6). 

The  upper  timberUne  receded  by  150-300  m  compared 
to  its  position  during  the  postglacial  thermal  optimvmi. 
That  timberline  depression  was  partly  due  to  a  general 
cooling.  However,  at  least  since  the  Middle  Ages  the  ef- 
fects of  the  deterioration  of  the  climate  on  subalpine  tree 
growth  have  been  compounded  or  even  superimposed  by 
the  human  influences  mentioned  earlier. 

COMPOSITION  AND  STRUCTURE 

As  a  consequence  of  these  activities  not  only  did  the  up- 
per timberline  become  lower  by  about  150-300  m,  but  also 
species  composition  and  structures  of  the  moimtain  for- 
ests changed  considerably  (Auer  1947;  Campell  1944; 
Holtmeier  1965,  1967).  Larch  (Larix  decidua)  and  cembra 
pine  {Pinus  cembra)  are  the  most  important  tree  species 
in  the  Upper  Engadine  forests.  These  forests  belong  to 
the  so-called  silicate  type  (Larici  Cembretum,  Ellenberg 
1978;  Larici-Pinetum  cembrae,  Oberdorfer  1970)  of  larch- 
cembra  pine  forest,  which  is  common  in  the  crystalline 
Central  Alps.  The  present  distribution  pattern  of  both 
species  in  the  Upper  Engadine  can  be  described  as  fol- 
lows: Cembra  pine  is  mainly  spread  on  steep  and  inacces- 
sible slopes  and  north  exposures,  whereas  larch  domi- 
nates on  easily  accessible  locations  exposed  to  the  south. 

The  influence  of  the  local  topography  on  the  distribu- 
tion pattern  is  most  pronovmced  in  the  trough-shaped 
Bever  Valley  and  Bernina  Valley  (fig.  7).  Pure  larch 
stands  are  confined  to  the  relatively  gentle  slope  areas 
(talus  cones,  screes,  boulder  fans,  etc.)  that  extend  below 
the  steep  and  rocky  trough  walls,  while  cembra  pine  pre- 
vails on  trough  walls  and  also  forms  the  upper  timberline. 
This  pattern  is  most  conspicuous  on  southern  exposures. 
On  north-exposed  slopes  cembra  pine  also  occurs  on  their 
lower  parts,  mixed  vnth  larch. 


This  distribution  of  larch  and  cembra  pine  is  mainly 
caused  by  human  influences,  especially  by  pastoral  use. 
Since  cembra  pine  impedes  the  growth  of  grass  and  herbs 
on  the  forest  floor,  while  larch  does  not,  cembra  pine  was 
cleared  on  slopes  suitable  for  grazing  and  restricted  to 
steep,  inaccessible  locations.  Moreover,  the  wood  of  cem- 
bra pine  was  used  for  many  purposes  such  as  paneling 
and  making  furniture,  milk  vessels,  or  carvings.  Cembra 
pine  also  is  more  susceptible  to  forest  fires  than  larch, 
which  is  protected  by  a  thick  corklike  bark.  As  indicated 
by  charcoal -rich  soil  horizons,  forest  fires  frequently 
occvured  in  the  past.  Probably,  many  of  them  were 
caused  by  shepherds  when  they  burned  alpine  pastures 
to  remove  "weeds"  such  as  Rhododendron  ferrugineum 
and  Juniperus  nana,  and  fires  ran  out  of  control.  Thus, 
for  many  reasons,  larch  co;ild  spread  at  the  cost  of  cembra 
pine,  as  was  also  the  case  in  many  other  central-alpine 
valleys.  At  many  places  the  high-elevation  forest  is  over- 
mature and  very  sensitive  to  environmental  impacts  such 
as  windthrow  and  snow  slides. 

CURRENT  CHANGES 

Due  to  modem  change  of  economic  structure — tourism 
has  become  the  main  base  of  the  economy  in  the 
Engadine — grazing  pressure  on  the  forests  declined,  and 
cembra  pine  is  invading  the  former  pasture  forest.  Larch, 
which  is  a  light-demanding  species,  is  going  to  be  out- 
competed  by  cembra  pine,  and  natural  succession  ft-om 
larch  to  cembra  pine  forests,  which  had  been  interrupted 
by  human  disturbances  for  some  hundred  years,  is  re- 
vived (Holtmeier  1967,  1990,  1993).  Larch  will  gradually 
be  replaced  by  cembra  pine,  if  sites  are  not  specially  man- 
aged by  exposing  mineral  soil  or  selective  cutting  of 
cembra  pine. 

Locally,  however,  invasion  of  cembra  pine  in  the  imder- 
story  is  hampered  by  mass  outbreaks  of  the  larch  bud 
moth  (Zeiraphera  diniana).  After  having  defoliated  the 
larch  crowns,  the  caterpillars  will  feed  on  the  cembra 
pines  in  the  understory  and,  together  with  secondary 
parasites,  cause  severe  damage  to  them.  Young  cembra 
pines  become  crippled  or  killed. 

In  forest  gaps  smaller  than  the  height  of  the  surround- 
ing trees,  accvunulation  of  snow  and  length  of  v^dnter  snow 
cover  may  critically  increase  (fig.  8),  and  consequently 
seedlings  and  saplings  of  cembra  pine  may  be  seriously 
damaged  by  snow  fungus  {Phacidium  infestans). 

Cembra  pine  also  is  invading  abandoned  or  only  rarely 
used  alpine  pastures.  Seeds  are  dispersed  by  the  nut- 
cracker {Nucifraga  caryocatactes),  occasionally  far  beyond 
the  present  forest  limit  (Holtmeier  1966,  1974;  Mattes 
1978,  1982).  Larch  seeds,  however,  are  prevented  from 
getting  into  a  suitable  seedbed  by  the  dense  alpine  vegeta- 
tion. Thus,  it  is  cembra  pine  that  is  the  pioneer  species 
on  former  pasture  areas  (fig.  9),  while  larch  only  occurs 
where  bare  mineral  soil  is  exposed.  Survival  of  the  invad- 
ing trees  mainly  depends  on  microclimates  and  resulting 
site  conditions.  They  considerably  deteriorated  when  the 
forest  was  removed  in  the  past.  Favorable  effects  of  the 
general  warming  on  tree  growth  are  superimposed  by 
unsuitable  local  site  conditions  (fig.  10).  Consequently, 
natural  reforestation  follows  the  favorable  sites  while 
other  sites  will  remain  treeless  probably  for  a  long  time. 


11 


Figure  7 — Distribution  of  tree  species  in  the  Bernina  Valley  (top)  and 
Bever  Valley  (bottom).  Both  maps  have  been  redrawn  from  Holtmeier 
(1967). 


12 


Figure  8 — Big  snow  masses  are  accumulated 
in  the  glade.  Incoming  solar  radiation  is  inter- 
cepted by  the  trees,  while  energy  loss  by  out- 
going long-wave  radiation  is  almost  unimpeded 
Due  to  these  conditions  snowmelt  is  consider- 
ably delayed,  and  young  cembra  pines  may  be 
infected  by  snow  fungus. 


As  to  successful  germination  and  survival  of  seedlings, 
seed  dispersal  by  the  nutcracker  is  much  more  effective 
than  wind-mediated  transport  because  many  seeds  are 
cached  at  sites  relatively  favorable  to  germination  and 
seedling  growth  (not  too  long  covered  with  snow).  On  the 
other  hand,  the  seed  caches  usually  provide  sufficient  soil 
moisture  because  they  are  2-4  cm  under  the  surface,  thus 
being  protected  from  direct  insolation.  Finally,  seeds  col- 
lected by  the  nutcracker  are  characterized  by  a  relatively 
high  germination  capacity  (Mattes  1978,  1982),  and  there 
is  almost  no  loss  of  cached  seeds  by  seed-eating  predators. 


Figure  9 — Cembra  pine  invading  an  abandoned 
alpine  pasture  (about  2,250  m)  on  the  northwest- 
exposed  slope  of  the  main  valley.  Invasion  fol- 
lows mainly  convex  topography,  while  sites 
covered  too  long  with  snow  and  characterized  by 
snow  creep  and  avalanches  remain  treeless. 


However,  due  to  high  density  of  seedling  clusters  some 
of  the  trees  will  become  victims  of  root  competition.  More- 
over, snow  accumulation  and  duration  of  the  snow  cover 
may  gradually  be  increased  by  the  influence  of  the  grow- 
ing trees  on  windflow  near  the  surface.  At  such  sites, 
yoimg  cembra  pines  covered  too  long  by  snow  may  be 
killed  by  snow  fungi  such  as  Phacidium  infestans 
(Holtmeier  1967,  1974,  1990  ). 

Natural  invasion  by  trees  (cembra  pine,  larch,  vallows, 
rowan,  and  also  spruce)  can  be  observed  in  front  of  and  on 
the  lateral  moraines  of  the  Morteratsch  Glacier  and  Roseg 
Glacier  (Holtmeier  1965,  1974,  1990;  fig.  11).  These  gla- 
ciers have  been  retreating  more  or  less  continuously  by 
about  30  m  per  year  since  the  beginning  of  our  century. 
Larch  alreadj  occurs  about  20  years  after  glacier  retreat, 
while  cembra  pine  ynW  follow  later.  Then,  however, 
cembra  pine  invades  these  areas  much  more  rapidly  than 
larch.  There  is  some  evidence  that  areas  which  became 
ice-free  about  50  to  70  years  ago  were  more  rapidly  in- 
vaded by  cembra  pine  than  those  from  where  the  glacier 
retreated  more  than  100  years  ago.  That  might  be  ex- 
plained by  mesoclimatic  conditions  improving  due  to  gla- 
cier retreat. 


7 


1.  Situation  during  the  postglacial  thermal  optimum 


forest  1  

 ecotone  1  alpine 

-  2100 

-  2100  ijpr 

forest  climate  replaced  by  the  very  locally 
changing  pattern  of  microclimates  ,  similar 
to  that  above  the  previous  tinfterline 

2.  Situation  at 

present 

Figure  10 — Change  of  microclimatic  conditions  in  the 
upper  Subalpine  after  the  uppermost  forest  was  re- 
moved by  humans. 


13 


1.5  km   1 


^Jk  cembrapine  ^^^i   larch  ''iV'    alpine  vegetation 

Figure  11 — Invasion  of  the  forefield  and  lateral  moraine  of  the  retreating  Morteratsch  Glacier  by  cembra 
pine  and  larch  (schematically). 


At  places,  however,  distribution  patterns  of  invading 
larches  and  cembra  pines  very  likely  depend  on  soil  mois- 
ture available,  as  can  be  observed  in  the  forefield  of 
Morteratsch  Glacier,  for  example  (fig.  11).  Obviously, 
relatively  dry  sites — such  as  the  well-drained  upper  part 
of  the  lateral  moraines  or  roche  moutonnee  areas — can  be 
invaded  by  cembra  pine  more  easily  than  by  larch,  be- 
cause seeds  of  cembra  pine  are  cached  by  the  nutcracker 
in  small  cracks  or  fine-soil  pockets  between  the  boulders 
where  sufficient  moisture  is  available  for  germination  and 
seedling  growth.  On  the  other  hand,  it  depends  totally  on 
chance  whether  wind-mediated  larch  seeds  will  get  into 
such  a  seedbed.  At  the  valley  bottom,  conditions  appear 
to  be  more  favorable  for  invasion  by  larch  because  of 
great  numbers  of  open  patches  of  bare  mineral  soil  cre- 
ated by  extensive  accumulation  of  fine  material  between 
the  boiilders.  At  such  sites  soil  moisture  is  usually  avail- 
able, and  there  is  a  better  chance  for  larch  seeds  to  reach 
a  suitable  seedbed. 


Spruce  (Picea  abies)  also  occurs  within  cembra  pine- 
larch  forest.  Natural  stands,  however,  are  confined  to  the 
northwest-exposed  slope  of  the  main  valley  between  the 
village  of  Sils  and  the  small  Lake  of  Staz  (fig.  12).  This 
area  is  characterized  by  increased  humidity  due  to  a  cloud 
belt  that  usually  appears  when  the  Upper  Engadine  is  in- 
fluenced by  moisture-carrying  air  currents  from  the 
southwest  (Holtmeier  1965,  1966,  1967).  The  upper  dis- 
tribution limit  of  spruce  is  located  at  about  2,000  m.  Al- 
though spruce  is  missing  between  Maloja  Pass  (1,815  m) 
and  Sils,  there  is  some  evidence  that  this  species  immi- 
grated from  the  south  across  Maloja  Pass  in  6000  to  5500 
B.C.  (Campell  1944;  Kleiber  1974).  The  gap  can  probably 
be  attributed  to  human  influences.  On  the  opposite  slope 
spruce  does  not  occur,  except  for  planted  trees  (Schlatter 
1935).  That  might  be  explained  by  lower  humidity  and 
lower  soil  moisture  on  that  slope  exposed  to  the  southeast. 
Anyway,  humans  also  could  have  removed  shade-giving 
spruce  (Holtmeier  1967).  After  deglaciation,  cembra  pine 
invaded  the  Upper  Engadine  earlier  than  spruce  (Kleiber 


Figure  12 — Distribution  of  spruce  {Picea  abies)  on  the  northwest-exposed  slope  of  the  main  valley  (from 
Holtmeier  1967,  modified). 


14 


Figure  13 — Arboreous  and  dwarfed  mountain  pine  in 
bog  near  Maloja  Pass. 


1974).  Otherwise  it  would  have  been  outcompeted  by  the 
latter  species — at  least  at  lower  elevations  (Heitz  1975; 
Mattes  1978). 

KRUMMHOLZ  SPECIES 

Dwarf  pine  {Pinus  mugo)  and  green  alder  (Alnus  viri- 
dis)  form  dense  and  locally  extended  krummholz  stands. 
The  krummholz  growth  of  both  these  species  is  genetically 
predetermined  and  not  climatically  induced.  That  makes 
them  different  from  climatically  trimmed,  flagged,  and 
matlike  "krummholz"  that  forms  the  forest-alpine  timdra 
ecotone  in  the  Rocky  Moimtains,  for  example  (Holtmeier 
1973,  1974,  1981). 

Moreover,  there  is  no  contiguous  Pinus  mugo  krumm- 
holz belt  above  the  high-stemmed  forest  as  is  peculiar  to 
the  northern  Alps.  In  the  Upper  Engadine,  both  Pinus 
mugo  and  Alnus  viridis  occur  mainly  on  slopes  endan- 
gered by  avalanches.  Their  elasticity  enables  them  to 
resist  avalanches  and  snow  slides  better  than  upright- 
growing  tree  species,  which  normally  will  be  eliminated 
at  such  sites.  Some  slopes,  where  the  original  forest  had 
been  destroyed  by  avalanches  and  fire,  were  completely 


afforested  with  moimtain  pine.  Moimtain  pine  is  also 
common  in  bogs  where  it  grows  on  slightly  convex  topog- 
raphy (mostly  rocky  outcrops).  At  such  sites  the  arbore- 
ous growth  form  also  occiu-s,  as  is  the  case  on  Maloja  Pass 
or  in  the  Lake  Staz  area,  for  example  (fig.  13). 

In  general,  Pinus  mugo  prevails  on  limestone  and  dolo- 
mite, while  Alnus  viridis  is  common  mainly  on  silicate 
material.  In  view  of  this  distribution,  both  species  are 
said  to  be  vicarious  plants.  In  the  Upper  Engadine,  how- 
ever, the  availability  of  soil  moistiu^e  seems  to  be  the  fac- 
tor controlling  the  distribution  pattern.  There,  Pinus 
mugo  is  more  common  on  southern  exposures,  whereas 
Alnus  viridis  is  mainly  spread  on  north-facing  slopes  pro- 
viding sufficient  soil  moistiire  and  fi'esh  soils  rich  in  nutri- 
ents. On  svmny  slopes,  however,  Alnus  viridis  is  confined 
to  moist  microsites,  as  alongside  rivlets,  for  example.  On 
northern  exposures  Pinus  mugo  successfully  competes 
with  Alnus  viridis  at  sites  relatively  dry  and  poor  in  nu- 
trients (fig.  14). 

Birch  {Betula  pubescens)  is  another  species  that  is  usu- 
ally confined  to  avalanche  tracks  and  bogs.  Birch  is  able 
to  recover  fi-om  breakage  by  thriving  basal  shoots.  There- 
fore, multistemmed  growth  is  quite  common  to  this  spe- 
cies (fig.  15). 

At  present,  the  Upper  Engadine  has  a  higher  forest 
cover  compared  to  150  years  ago  and  earlier.  Since  then 
many  afforestations  were  carried  out  to  close  the  gaps 
caused  by  historical  hiunan  impact.  Also  "exotic"  species 
were  used — such  as  Picea  pungens,  Picea  engelmannii,  Pi- 
nus sibirica,  and  others  (Schlatter  1935).  It  has  become 
evident,  however,  that  in  this  area  the  native  Pinus  cem- 
bra  and  Larix  decidua  are  the  most  successful  species. 

CONCLUDING  REMARKS 

The  Upper  Engadine  belongs  to  the  most  important 
natural  ranges  of  cembra  pine-larch  forests  in  the  Alps. 
However,  present  distribution,  composition,  and  struc- 
tures of  these  forests  can  only  be  explained  by  permanent 
utilization  that  began  in  prehistoric  time.  That  makes 
these  forests  much  different  from  almost  imtouched  high- 
altitude  forests  in  remote  areas  of  the  Rocky  Mountains, 
for  example. 

Since  tourism  has  become  the  main  economic  base,  while 
agriculture  and  pastoral  land  use  declined,  utilization  of 
the  forest  also  changed.  In  our  day  it  is  not  timber  produc- 
tion but  establishment  and  maintenance  of  a  vigorous 
shelter  forest  (protection  from  avalanches,  soil  erosion, 
torrential  washes,  etc.)  that  is  the  main  objective  of  forest 
management.  Locally,  grazing  of  the  high-altitude  forest 
by  cattle  and  game  may  run  counter  to  the  objective  and 
should  be  reduced  as  much  as  possible. 

However,  the  forest  also  is  of  great  importance  for  tour- 
ism. Thus,  larch,  which  has  a  high  esthetic  value  because 
of  its  bright  autvmin  colors,  is  favored  by  special  manage- 
ment practices  interrupting  natural  succession.  On  the 
other  hand,  lower  elevation  forest  is  still  used  for  produc- 
tion of  merchantable  timber.  Swiss  stone  pine  is  used  in 
particular  for  paneling,  carvings,  and  making  furniture  of 
typical  local  design. 


15 


PINUS  MUCO 

LARIX  DECIDUA 

9  <* 

PINUS  CEMBRA 


Figure  14 — Distribution  of  Pinus  mugo  and  AInus  viridis  on  the  northeast-exposed  slope  of  the  Bernina  vaiiey. 
Pinus  mugo  is  confined  to  relatively  dry  convex  topography,  while  AInus  viridis  grows  on  the  moist  sites  (from 
Holtmeier  1967). 


REFERENCES 

Auer,  Christian.  1947.  Untersuchungen  iiber  die  natiir- 
liche  Verjiingung  der  Larche  im  Arven-Larchenwald 
des  Oberengadins.  Mitteilungen  der  Schweizerischen 
Anstalt  fiir  das  forstliche  Versuchswesen.  25:  3-140. 

Campell,  Eduard.  1944.  Der  Wald  des  Oberengadins  im 
Wandel  der  Zeiten.  Festschrift  124.  Jahresversamm- 
lung  der  Schweizerischen  Naturforschenden 
Gesellschaft:  93-102. 

Conrad,  H.  1940.  Beitrag  zur  Frage  der  urgeschichtHchen 
Besiedlung  des  Engadins.  Jahresbericht  Historisch- 
Antiquarische  Gesellschaft  Graubixndens.  70. 
Jahresbericht:  1-40. 


Figure  15 — Birch  {Betula  pubescens)  growing 
in  an  avalanche  track. 


Ellenberg,  Heinz.  1978.  Vegetation  Mitteleuropas  mit  den 
Alpen  in  okologischer  Sicht.  2d  ed.  Stuttgart:  Verlag 
Eugen  Ulmer.  981  p. 

Heitz,  Christian.  1975.  Vegetationsentwicklung  und 
Waldgrenzschwankungen  des  Spatglazials  iind  Post- 
glazials  im  Oberhalbstein  (Graubiinden/Schweiz). 
Biindnerwald.  28(2):  45-52. 

Holtmeier,  Friedrich-Karl.  1965.  Die  Waldgrenze  im 
Oberengadin  in  ihrer  physiognomischen  und  okolo- 
gischen  DifFerenzierung.  University  of  Bonn,  FRG. 
163  p.  Dissertation. 

Holtmeier,  Friedrich-Karl.  1966a.  Die  Malojaschlange 
und  die  Verbreitung  der  Fichte.  Beobachtungen  zur 
Klimaokologie  des  Oberengadins.  Wetter  und  Leben. 
18:  105-108. 

Holtmeier,  Friedrich-Karl.  1966b.  Die  okologische  Funk- 
tion  des  Tannenhahers  im  Zirben-Larchenwald  und  an 
der  Waldgrenze  im  Oberengadin.  Journal  fur  Omi- 
thologie.  4:  253-260. 

Holtmeier,  Friedrich-Karl.  1967.  Die  Verbreitung  der 
Holzarten  im  Oberengadin  unter  dem  Einflui3  des 
Menschen  und  des  Lokalklimas.  Erdkunde.  21:  249-258. 

Holtmeier,  Friedrich-Karl.  1973.  Geoecological  aspects  of 
timberlines  in  Northern  and  Central  Europe.  Arctic  and 
Alpine  Research.  5(3):  45-54. 

Holtmeier,  Friedrich-Karl.  1974.  Geookologische  Beobach- 
tungen und  Studien  an  der  subarktischen  und  alpinen 
Waldgrenze  in  vergleichender  Sicht  (nordliches  Fenno- 
skandien/Zentralalpen).  Erdwissenschaftliche  For- 
schung.  8.  130  p. 

Holtmeier,  Friedrich-Karl.  1981.  What  does  the  term 
"Krummholz"  really  mean?  Observations  with  special 
reference  to  the  Alps  and  the  Colorado  Front  Range. 
Mountain  Research  and  Development.  1(3/4):  253-260. 

Holtmeier,  Friedrich-Karl.  1990.  Disturbance  and  man- 
agement in  larch-cembra  pine  forests  in  Europe.  In: 


16 


Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps.  Pro- 
ceedings— symposium  on  whitebark  pine  ecoystems: 
ecology  and  management  of  a  high-momitain  resoiirce. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agricultvire,  Forest  Service,  Intermoxantain  Research 
Station:  25-36. 

Holtmeier,  Friedrich-Karl.  1993.  European  larch  in 
Middle  Europe.  In:  Proceedings — symposium  on  ecology 
and  management  of  larix  forests:  a  look  ahead.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain  Research  Station.  [In  preparation]. 

Kleiber,  Helga.  1974.  Pollenanalytische  Untersuchvmgen 
zum  Eisriickgang  und  zur  Vegetationsgeschichte  im 
Oberengadin  I.  Botanisches  Jahrbuch  fiir  Systematik. 
94(1):  1-53. 


Mattes,  Hermann.  1978.  Der  Tannenhaher  im  Engadin — 

Studien  zu  seiner  okologischen  Funktion  im  Arvenwald. 

Miinstersche  Geographische  Arbeiten.  2.  87  p. 
Mattes,  Hermann.  1982.  Die  Lebensgemeinschaft  von 

Tannenhaher  und  Arve.  Eidgenossische  Anstalt  fur  das 

forstliche  Versuchswesen.  Berichte.  24.  74  p. 
Oberdorfer,  Erich.  1970.  Pflanzensoziologische  Exkur- 

sionsflora  fiir  Siiddeutschland.  3d  ed.  Stuttgart:  Verlag 

Eugen  Ulmer.  987  p. 
Schlatter,  A.  J.  1935.  Die  Aufforstungen  und  Verbau- 

xmgen  des  Oberengadins  in  den  Jahren  1875-1934. 

Schweizerische  Zeitschrift  fur  Forstwesen.  86:  309-328. 


17 


Evolution  and 
Taxonomy 


International  Workshop 
St.  Moritz  1 992 

S   


18 


GENETIC  DIFFERENTIATION  AND 
PHYLOGENY  OF  STONE  PINE 
SPECIES  BASED  ON  ISOZYME  LOCI 

Konstantin  V.  Krutovskii 
Dmitri  V.  Politov 
Yuri  P.  Altukhov 


Abstract— F-statistics  analysis  based  on  isozyme  loci  data  has 
revealed  comparatively  low  levels  of  intraspecific  genetic  differen- 
tiation among  natural  populations  of  Eurasian  stone  pine  species: 
Siberian  stone  pine  {Pinus  sibirica  Du  Tour),  Korean  stone  pine 
{Pinus  koraiensis  Siebold  et  Zucc),  and  Japanese  (dwarf  Siberian 
or  mountain)  stone  pine  {Pinus  pumila  [Pall.]  Kegel).  Only  about 
2  to  4  percent  of  the  total  isozyme  gene  diversity  was  due  to  inter- 
population  variation  (FgT  =  0.021-0.040),  and  the  overwhelming 
part  of  the  total  variation,  over  96  percent,  belonged  to  intrapopu- 
lation  variation.  These  data  are  discussed  in  relation  to  distin- 
guishing features  of  conifer  biology,  in  general,  and  to  a  specific 
mode  of  stone  pine  seed  distribution  by  nutcrackers,  in  particular. 
Cluster  analysis  of  more-studied  Siberian  stone  pine  populations 
has  shown  that  genetic  distances  based  on  isozyme  allele  frequen- 
cies reflect  geographical  distribution  of  these  populations  in  spite 
of  the  generally  low  values  of  genetic  distances  (D  =  0.001-0.031, 
on  the  average  of  0.006). 

The  authors  have  studied  phylogenetic  relationships  between  all 
stone  pine  species  (subsection  Cembrae),  including  one  European 
stone  pine  population— Swiss  stone  pine  {Pinus  cembra  L.)  and  one 
population  of  the  only  representative  of  stone  pine  species  in  North 
America— whitebark  pine  {Pinus  albicaulis  Engelm.),  using  genetic 
distances  based  on  16  isozjmie  loci,  whose  genetic  control  has  been 
resolved  for  all  stone  pine  species.  The  results  have  confirmed 
stone  pine  {Cembrae)  species  to  be  a  compact  group  of  very  closely 
related  and  genetically  similar  species,  having  monophyletic  ori- 
gin, supposedly  from  ancient  Siberian  stone  pine.  The  smallest 
distance  (D  =  0.065)  has  been  found  between  Pinus  cembra  and 
Pinus  sibirica  that  corresponded,  mostly,  to  their  subspecies  taxo- 
nomic  status.  Other  interspecific  distances  were  in  good  agree- 
ment with  the  species  status  (D  =  0.121-0.268). 

As  to  the  controversy  on  the  taxonomy  of  whitebark  pine,  this 
species  undoubtedly  belongs  to  subsection  Cembrae  (D  =  0.121- 
0.256).  The  results  of  phylogenetic  study  have  been  confirmed  by 
the  recently  obtained  data  of  cpDNA  restriction  fragment  analyses. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  ICnowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Konstantin  V.  Krutovskii  is  Senior  Research  Scientist,  Laboratory  of 
Population  Genetics,  N.I.  Vavilov  Institute  of  General  Genetics,  Russian 
Academy  of  Sciences,  GSP-1  Moscow  117809  B-333,  Russia  (current  ad- 
dress until  1995:  Department  of  Forest  Genetics  and  Tree  Improvement, 
Georg-August-University,  Biisgenweg  2,  Gottingen,  37077,  Germany); 
Dmitri  V.  Politov  is  Research  Scientist,  same  laboratory;  Yuri  P.  Altukhov 
is  Professor,  Member  of  Russian  Academy  of  Sciences,  Head  of  the  Labora- 
tory of  Population  Genetics,  Director  of  the  Institute  of  General  Genetics. 


Five  stone  pine  species  are  traditionally  united  in  sub- 
section Cembrae  of  section  Strobus,  subgenus  Strobus, 
genus  Pinus  (Critchfield  and  Little  1966;  Shaw  1914). 
European  or  Swiss  stone  pine  {Pinus  cembra)  is  distrib- 
uted in  the  Alps  and  Carpathians  and  has  the  smallest 
area  among  stone  pine  species.  An  east  Asian  stone  pine 
species,  Korean  pine  (P.  koraiensis),  is  widely  distributed 
in  eastern  Russia  (Russian  "Far  East"),  Manchuria,  and 
Japan,  but  mostly  in  Korea.  Two  other  stone  pine  species 
are:  Japanese  stone  pine  {P.  pumila),  widely  distributed 
in  northeastern  Siberia  extending  to  eastern  Asia,  as  far 
as  central  Japan;  and  Siberian  stone  pine  (P.  sibirica), 
widely  distributed,  mostly  in  Siberia  extending  to  northern 
Mongolia  and  to  the  eastern  border  of  Europe.  Whitebark 
pine  (P.  albicaulis)  is  widely  distributed  in  western  North 
America,  extending  from  British  Coliunbia  to  the  southern 
Sierra  Nevada,  and  represents  the  only  stone  pine  species 
growing  outside  Eurasia. 

All  these  species  are  of  great  interest  for  different  kinds 
of  investigations,  but  some  main  aspects  make  these  species 
very  attractive  from  evolutionary  and  population  genetics 
points  of  view  and  explain  the  necessity  of  their  genetic 
study. 

First,  they  occupy  a  vast  territory  in  Eurasia  and  North 
America  and  most  of  them  belong  to  the  main  forest-forming 
species.  Thus,  they  have  great  ecological  and  economic  sig- 
nificance, especially  in  Eurasia.  They  are  subjected  to  inten- 
sive harvesting  and,  consequently,  need  special  breeding 
and  gene  conservation  programs  based  on  genetic  data. 

Second,  stone  pine  species  consist  of  different  kinds  of  pop- 
ulations. For  example,  Swiss  stone  pine  includes  mostly 
small,  isolated  populations  and  occupies  a  relatively  small 
territory.  On  the  other  hand,  Siberian  stone  pine  includes 
huge,  continuous  populations  occupying  a  very  leirge  terri- 
tory in  Siberia.  Some  of  the  stone  pine  species  are  com- 
pletely isolated  from  others:  Swiss  stone  pine,  growing  in 
Europe,  and  whitebark  pine,  growing  in  North  America, 
stand  apart  from  Asian  stone  pine  species.  At  the  same 
time,  Siberian,  Korean,  and  Japanese  stone  pines  have 
sympatric  zones.  Thus,  from  evolutionary  and  population 
genetics  points  of  view  and  for  gene  conservation  programs 
it  would  be  interesting  and  important  to  study  genetic  dif- 
ferentiation, gene  dispersal  patterns,  and  mating  system 
and  population  genetic  structure  of  these  species  and  com- 
pare them  with  each  other. 

Third,  in  spite  of  the  fact  that  stone  pines  have  a  wide 
geographical  distribution  and  some  morphological  and 
ecological  differences,  they  are  grouped  into  a  separate 
subsection  Cembrae,  because  they  have  some  diagnostic 


19 


morphological  traits  that  distinguish  them  from  other 
Strobus  species.  Most  significant  of  these  are  wingless  large 
seeds  and  indehiscent  cones  that  remain  closed  in  matu- 
rity. These  traits  are  supposed  to  be  the  result  of  long  co- 
evolution  of  stone  pines  and  nutcrackers  (Nucifraga  ssp.), 
who  play  a  crucial  role  in  natural  reproduction  and  dis- 
tribution of  stone  pine  species  (Lanner  1980,  1982,  1990; 
Tomback  1983;  Tomback  and  Linhart  1990;  Tomback  and 
others  1990).  Nevertheless,  taxonomy  of  these  species  is 
still  being  discussed.  Some  specialists  consider  Swiss  stone 
pine  a  subspecies  of  Siberian  stone  pine  (or  vice  versa)  and 
some  think  that  whitebark  pine  should  be  included  into 
subsection  Strobi  because  of  its  morphological  and  ecologi- 
cal resemblance  to  some  Strobi  species  (Critchfield  1986; 
Millar  and  Kinloch  1991).  Taking  all  this  into  account,  the 
use  of  genetic  markers  for  study  of  phylogenetic  relation- 
ships between  stone  pine  species  would  be  very  appropriate 
and  fruitful. 

Thus,  the  main  tasks  of  our  research  were  estimation  of 
genetic  differentiation  and  subdivision  of  stone  pine  popula- 
tions and  study  of  phylogenetic  relationships  between  stone 
pine  species  using  isozyme  loci  as  genetic  markers. 

The  advantages  and  disadvantages  of  using  isozymes  in 
plant  systematics  and  phylogeny  have  been  fairly  thoroughly 
discussed  in  scientific  literature  (see  for  example,  Crawford 
1983;  Gottheb  1977;  Strauss  and  others  1992).  We  would 
only  like  to  note  here  such  taxonomically  most  important 
advantages  as,  first,  the  possibility  of  using  the  same  trait 
(isozyme  electrophoretic  pattern  encoded  by  homologous 
loci)  to  classify  taxons  of  almost  any  phylogenetic  level,  both 
related  and  unrelated,  and  second,  the  possibility  of  quan- 
titative estimation  of  divergence  degree.  One  of  the  prob- 
lems of  traditional  systematics  is  associated  with  the  objec- 
tive choice  of  diagnostic  traits,  but  initial  selection  of  such 
traits  usually  leads  to  their  preliminary  "weighting"  (evalu- 
ation) and  inevitably  produces  subjectivity.  It  is,  therefore, 
extremely  important  to  use,  above  all,  such  traits  in  taxo- 
nomic  studies  as  can  be  taken  for  analysis  without  "weigh- 
ing," more  or  less  arbitrarily,  and  about  whose  variability 
nothing  has  been  known  a  priori.  Molecular  genetic  mark- 
ers (isozyme  loci  and  DNA  markers)  are  one  of  the  best, 
from  these  points  of  view. 

The  study  presented  here  is  the  first  one  devoted  to  study 
of  phylogeny  of  the  whole  subsection  Cembrae  using  isozyme 
markers  and  is  only  a  part  of  long-term  studies  on  popula- 
tion genetics  of  stone  pine  species  carried  out  in  the  Labora- 
tory of  Population  Genetics  of  N.  I.  Vavilov  Institute  of  Gen- 
eral Genetics  (Russian  Academy  of  Sciences,  Moscow,  Russia) 
on  the  basis  of  isozyme  analysis.  More  details  on  inheritance 
and  linkage  of  isozjnne  loci,  mating  system,  population  ge- 
netic structure,  and  other  aspects  of  stone  pine  population 
genetics  can  be  found  in  our  previously  published  papers 
(Krutovskii  and  Politov  1992;  Krutovskii  and  others  1987, 
1988,  1989,  1990;  PoHtov  1989;  Politov  and  Krutovskii  1990; 
Politov  and  others  1989,  1992)  and  in  the  accompanying  pa- 
per included  in  these  proceedings  by  Politov  and  Krutovskii. 

MATERIAL  AND  SAMPLING  SITES 

Seventeen  single-tree  and  two  bulk  seed  samples  from 
natural  populations  have  been  used  for  isozyme  analysis  in 
our  study.  Eleven  localities  of  Siberian  stone  pine,  three 


localities  each  of  Korean  and  Japanese  stone  pines,  and 
one  locality  each  of  Swiss  stone  and  whitebark  pines  have 
been  studied  on  16-20  isoz3niie  loci  (table  1).  More  details 
about  seed  samples  and  characteristics  of  the  sampled 
natural  populations  can  be  found  in  our  other  publications 
(Krutovskii  and  Politov  1992;  Krutovskii  and  others  1988, 
1989,  1990;  Pohtov  1989;  Pohtov  and  Krutovskii  1993; 
Pohtov  and  others  1992). 

ELECTROPHORETIC  ANALYSIS 

Isozymes  of  seeds  have  been  studied  by  electrophoresis 
in  starch  gel.  Method  of  electrophoretic  analysis,  specimen 
preparation,  the  buffer  systems  used,  genetic  interpretation 
of  zymograms,  designations  of  allozymes,  alleles,  and  loci, 
and  inheritance  of  enzyme  systems  have  been  described 
elsewhere  (Krutovskii  and  others  1987;  Politov  1989).  Geno- 
types of  trees  have  been  determined  by  analyzing  segrega- 
tion of  allozyme  alleles  among  their  endosperms,  which  are 
haploid  in  conifers  and  genetically  identical  to  maternal 
tree  gametes.  Together  with  adult  tree  allele  frequencies, 
pollen  allele  frequencies  inferred  from  effective  pollen  gene 
pool  (paternal  alleles  of  the  analyzed  seed  embryos)  have 
been  used  for  estimation  of  genetic  distances.  These  fre- 
quencies of  successful  pollen  give  probably  more  represen- 
tative estimation  of  the  studied  populations'  gene  pools  due 
to  predominantly  outcrossing  pollination  among  stone  pine 
species  (Pohtov  and  Krutovskii  1990,  1993). 

In  total,  536  trees  fi-om  17  localities  (11-75  trees  each)  and 
two  bulk  seed  samples  of  five  stone  pine  species  have  been 
analyzed  on  16-20  isozyme  loci.  More  than  3,500  (7-20  per 
tree)  embryos  have  been  analyzed  on  a  smaller  amount  of 
isoz5Tne  loci  because  of  poor  resolution  of  some  enzyme  sys- 
tems in  embryo  tissue.  A  relatively  small  number  of  trees 
have  been  analyzed  in  some  populations  and  species,  but 
according  to  some  theoretical  (Nei  1987a,b)  and  experimen- 
tal investigations  (Shurkhal  and  others  1992),  the  number 
of  loci  used  in  interspecific  phylogenetic  analysis  is  more 
important  than  the  sample  size.  We  have  studied  phyloge- 
netic relationships  between  all  stone  pine  species  using  ge- 
netic distances  based  on  16  isozyme  loci,  Adh-1,  Dia-2,  Fe-2, 
Gdh,  (iot-1,  Got-2,  Got-3,  Idh,  Mdh-1,  Mdh-2,  Mdh-3,  Mdh-4, 
Pgi-2,  Pgm-1,  Pgm-2,  and  Skdh-1,  whose  allelic  variation  has 
been  unequivocally  determined  for  all  stone  pine  species. 
Such  a  number  of  loci  is  used  commonly  and  gives  quite 
reasonable  phylogenetic  trees  (Nei  1987a,b). 

DATA  ANALYSIS 

Determination  of  parameters  of  intra-  and  interpopula- 
tional  genetic  diversity,  estimation  of  genetic  differentia- 
tion and  genetic  distances,  and  clustering  and  construction 
of  dendrograms  for  stone  pine  populations  have  been  car- 
ried out  by  using  IBM  PC  version  1.7  of  the  computer  pro- 
gram BIOSYS-1  (Swofford  and  Selander  1981)  and  IBM 
PC-program  RESTSITE,  version  1.1  (Nei  and  Miller  1990). 

Using  our  data,  we  have  estimated  different  kinds  of  ge- 
netic distances  and  tried  different  methods  of  clustering. 
Methods  for  calculating  genetic  distances  and  methods  of 
clustering  differ  in  their  assumptions  concerning  homoge- 
neity of  evolutionary  rates  or  proportionality  to  the  time  of 
divergence.  Consequently,  topology  of  derived  phylogenetic 


20 


Table  1 — Name,  abbreviated  designation,  sample  size,  and  geographic  location  of  sampling  sites 


Locality 


Designation 


Number 
of  trees 


Region 


Siberian  Stone  Pine 


Filin  Klyuch 


Mutnaya  Rechka  M 

Maiyi  Kebezh  MK 

Sobach'ya  Rechka  SR 

Listvyanka  L 

Yailyu  YA 

Smokotino  S 

Zorkartsevo-1  Z-1 

Zorkartsevo-3  Z-3 

Turukhansk  TU 


Nyalino  NYA 


Ust'-Chornaya  UCH 


Grustnyi  GR 
Uttyveem  UT 
Kamenskoye  K 


Malokhekhtsirskoye  MKH 
Khor  KHO 
Sikhote-Alin'  SA 


Livingstone  Falls  LF 


53 


75 
15 
12 
41 
43 

43 
34 
34 

MOO 


Western  Sayan  Mountains  (Ermakovskii 
District,  Krasnoyarsk  Territory, 
eastern  Siberia) 


Altai  Mountains  (Gorno-Altaiskaya 

Autonomous  Region,  eastern  Siberia) 
Tomskaya  Region  (western  Siberia) 


'88 


North  of  Krasnoyarsk  Territory, 
(Turukhanskii  District,  North  of 
eastern  Siberia) 
Khanty-Mansi  Autonomous  District 
(Tyumenskaya  Region,  North  of 
western  Siberia) 

Swiss  Stone  Pine 

15  Eastern  Carpathians  (western  Ukraine) 

Japanese  Stone  Pine 

54  North  of  Kamchatka  Peninsula 

55 

55 


Korean  Stone  Pine 


30 
16 
11 

Whitebark  Pine 

52 


Khabarovsk  Territory 

(Russian  "Far  East") 

Primorskii  Territory  (Russian  "Far  East") 


Rocky  Mountains  Forest  Reserve, 
Bow-Crow  Forest  (southwestern 
Alberta,  Canada) 


'Number  of  analyzed  seeds  from  mixed  sample  obtained  from  large  number  (>100)  of  trees. 


trees  based  on  different  distances  and  methods  of  cluster- 
ing can  vary  significantly  for  the  same  data. 

Among  methods  of  clustering  that  assimie  homogeneity 
of  evolutionary  rates,  an  unweighted  pair  group  method 
using  arithmetic  averages,  UPGMA,  is  the  most  widely 
used  one  (Sneath  and  Sokal  1973).  Among  methods  that 
do  not  assume  homogeneity  of  evolutionary  rates,  the  dis- 
tance Wagner  procedure  (Farris  1972)  and  neighbor-joining 
method,  NJM  (Saitou  and  Nei  1987)  are  the  most  widely 
used  ones. 

The  most  common  genetic  distances  used  in  isozjnne 
studies  of  trees  are  Nei's  (Nei  1972,  1978),  Rogers'  (1972), 
and  CavalU-Sforza  and  Edwards'  (1967)  distances.  How- 
ever, Nei's  distance,  D,  along  with  advantages  has  some 
disadvantages,  the  main  one  being  that  it  is  not  metric 
because  it  does  not  obey  the  triangle  inequality  and,  conse- 
quently, cannot  be  used  in  the  distance  Wagner  procedure 
or  NJM.  Thus,  we  have  used  D  (Nei  1972)  only  in  UPGMA 
dendrograms. 


Taking  into  accoimt:  (1)  that  homogeneity  of  evolution- 
ary (or  allozyme  allele  substitution)  rates  can  usually  be 
safely  assumed  in  comparing  conspecific  populations  of 
widespread,  continuous  tree  species  with  high  gene  flow, 
like  stone  pine  populations;  (2)  shortage  of  space  for  pre- 
sentation of  all  obtained  dendrograms;  and  (3)  possibility 
for  wider  comparison  vnth  published  data,  we  will  present 
only  UPGMA  dendrograms  based  on  Nei's  distance,  D, 
for  analysis  of  intraspecies  stone  pine  population  relation- 
ships. However,  to  avoid  problems  connected  with  possible 
nonhomogeneity  of  evolutionary  rates  of  different  species, 
we  will  present  NJM  dendrograms  of  stone  pine  species 
based  on  Cavalli-Sforza  and  Edwards'  (1967)  chord  dis- 
tances, Dch,  in  addition  to  UPGMA  dendrograms  based  on 
D  and  D.^. 

Nei's  genetic  distances  matrix  has  also  been  used  for 
principal  coordinate  analysis  of  Siberian  stone  pine  popula- 
tions using  the  computer  program  NTSYS-pc  (Rohlf  1988), 


21 


INTRASPECIFIC  DIFFERENTIATION 

Levels  of  Genetic  Differentiation  Between 
Populations — Distribution  of  intraspecific  genetic  variation 
between  populations  of  stone  pines  has  been  analyzed  for 
three  species  studied  in  several  localities  using  F-statistics 
(Nei  1977).  Levels  of  genetic  differentiation  among  popula- 
tions were  comparatively  low — only  about  2-4  percent  of 
the  total  intraspecific  isozyme  gene  variation  was  due  to 
interpopulation  variation  (FgT  =  0.021-0.040),  and  the  over- 
whelming part  of  the  total  variation,  over  96  percent,  be- 
longed to  intrapopulation  variation  (table  2).  Loci  Fe-2, 
Mdh-3,  and  Skdh-1  of  Siberian  stone  pine,  Adh-1,  Adh-2, 
Gdh,  Got-2,  Mdh-2,  Mdh-3,  Mdh-4,  and  Skdh-2  of  Japanese 
stone  pine,  and  Mdh-4  of  Korean  stone  pine  make  a  most 
significant  contribution  in  intraspecific  differentiation  of 
these  species  (table  2).  The  geographic  variation  of  these 
loci  may  be  adaptive  and  should  be  studied  in  detail  in  fu- 
ture investigations. 

Nei's  genetic  distances  between  populations  were  also 
small,  D  =  0.001-0.031  and  0.012,  on  the  average  (table  3). 
These  values  correspond  perfectly  to  the  calculations  car- 
ried out  using  references  on  22  pine  species  (Altukhov  and 
others  1989).  The  genetic  distances  between  closely  located 
(<20  km)  populations  of  seven  pine  species  equaled  0.009, 
on  the  average,  and  between  widely  dispersed  populations 
of  15  pine  species  reviewed — D  =  0.036,  on  average. 

Low  levels  of  interpopulation  differentiation  of  allozyme 
loci  are  usual  for  conifers.  Such  typical  factors  for  most  coni- 
fer populations  as  outcrossing,  wind  pollination,  seed  dis- 
persal by  wind  (most  conifers)  or  by  birds  (stone  pines,  some 
white  pines,  and  pinyons),  wide  continuous  ranges,  high 
population  density,  and  effective  size  are  considered  to  re- 
duce the  influence  of  genetic  drift  and,  therefore,  decrease 


heterogeneity  of  allele  frequencies  and  interpopulation 
genetic  differentiation  (Hamrick  1983;  Hamrick  and  Godt 
1989;  Hamrick  and  Loveless  1986;  Hamrick  and  others 
1981;  Loveless  and  Hamrick  1984). 

We  have  also  simmiarized  data  on  interpopulation  genetic 
differentiation  of  28  conifer  species  of  eight  genera  estimated 
using  FgT  or  analogous  GgT  parameters  of  differentiation 
(Politov  and  others  1992).  In  spite  of  a,  generally,  compar- 
atively low  differentiation  among  populations,  we  have 
found  a  quite  pronounced  tendency  to  a  significantly  smaller 
differentiation  (Fst  =  0.02-0.08)  among  widespread  species 
with  continuous  population  ranges  than  among  species  with 
interrupted  ranges  and  small,  and  often  isolated,  popula- 
tions (Fg-r  =  0.11-0.13).  Moreover,  according  to  our  review, 
samples  from  the  same  population  (samples  fi'om  localities 
within  several  kilometers  radius)  revealed,  commonly,  an 
even  much  lower  level  of  differentiation  (Fg^  =  0.006-0.02). 

The  three  stone  pine  species  studied  belong  to  the  first 
group  of  species  with  mainly  continuous  population  range. 
Besides,  four  of  the  11  Siberian  stone  pine  samples,  F,  M, 
MK,  and  SR,  have  been  collected  within  20  km  radius  from 
obviously  the  same  population  of  Malyi  Kebezh  River  Basin 
(eastern  Siberia,  table  1);  and  the  other  two  samples  have 
been  taken  fi^om  two  stands  near  Zorkal'tsevo  village  (west- 
em  Siberia),  Z-1  and  Z-3,  which  are  separated  from  each 
other  by  less  than  3  km.  The  other  localities  are  rather  re- 
mote from  one  another,  but  all  of  them  represent  only  the 
central,  continuous  part  of  the  species  area.  There  are  no 
significant  barriers  to  gene  flow  in  this  area. 

The  low  differentiation  observed  in  our  study  may  also 
be  due  to  the  limited  number  of  samples  taken,  mostly, 
from  more  or  less  optimal  zones  of  the  stone  pines  area, 
impeding  the  diversifying  selection  action  and  promoting 
the  balancing  one  (Altukhov  1990). 


Table  2 — The  estimates  of  F-statistics  for  populations  of  three  stone  pine  species  and  for  all  five  species 


Locus 

Siberian  stone  pine 

Korean  stone  pine 

Japanese  stone  pine 

All 

Ft 

Ft 

Fst 

F,s 

Ft 

Fst 

Fst 

Adh-1 

-0.030 

-0.006 

0.023 

0.078 

0.120 

0.045 

0.306 

0.321  ' 

0.022 

'0.563 

Adh-2 

{') 

{') 

{') 

(') 

{') 

-.068 

-.046 

\020 

Dia-1 

-.032 

-.007 

.024 

(') 

(') 

{') 

n 

{') 

Dia-2 

-.060 

-.053 

.006 

{') 

{') 

(') 

-.038 

-.013 

\025 

'.028 

Fe-2 

-.061 

-.012 

\046 

-.067 

-.021 

.043 

-.105 

-.097 

.007 

'.190 

Gdh 

-.012 

-.001 

.010 

-.098 

-.098 

.000 

.084 

.140 

^061 

'.637 

Got-1 

-.013 

-.001 

.012 

{') 

{') 

{') 

{') 

{') 

{') 

.002 

Got-2 

-.012 

-.001 

.010 

{') 

e) 

-.050 

-.022 

\027 

'.016 

Got-3 

(') 

{') 

{') 

.238 

.241 

.004 

.014 

.015 

.001 

'.596 

Lap-2 

.015 

.025 

.009 

-.071 

-.023 

.045 

-.048 

-.047 

.001 

Lap-3 

.066 

.079 

.013 

(') 

{') 

{') 

-.064 

-.050 

.013 

Mdh-2 

.214 

.231 

.023 

-.251 

-.225 

.020 

-.004 

.015 

^019 

'.240 

Mdh-3 

-.049 

.000 

\047 

-.067 

-.021 

.043 

-.051 

.084 

M28 

'.118 

Mdh-4 

-.068 

-.055 

.012 

.017 

.104 

'.089 

-.021 

-.013 

\ooa 

'.362 

Pgi-2 

.114 

.123 

.010 

e) 

{') 

{') 

-.111 

-.107 

.003 

'.407 

Pgm-1 

-.074 

.051 

.022 

-.193 

-.160 

.028 

.177 

.189 

.016 

'.187 

Pgm-2 

{') 

(') 

{') 

{') 

?) 

-.014 

-.013 

.002 

'.046 

Skdh-1 

-.146 

-.118 

\025 

-.316 

-.194 

.092 

-.053 

-.051 

.002 

'.187 

Skdh-2 

{') 

{') 

{') 

{') 

{') 

.099 

.163 

'.071 

Mean 

-.043 

-.018 

.025 

-.077 

-.033 

.040 

-.002 

.020 

.021 

.396 

'This  locus  has  statistically  significant  heterogeneity  of  allele  frequencies  among  populations. 

n"he  genetic  base  of  isozyme  variation  of  this  locus  is  not  determined  or  allelic  variation  cannot  be  compared  with  other  species, 
^his  locus  is  monomorphic  in  this  species. 


22 


Table  3 — Genetic  distances,  D,  (Nei  1972)  for  five  stone  pine  species 


Sibarian 

Ivor  oan 

Species 

stone  pine 

stone  pine 

stone  pine 

Stone  pine 

Siberian  stone  pine 

11 

0.006 
(0.002-0.016) 

Swii^^  stonp  oinp 

"1 

0  065 
(0.055-0.077) 

Jaoanese  stone  oine 

3 

0.202 
(0.176-0.253) 

0.233 
(0.215-0.265) 

0  010 
(0.003-0.014) 

Korean  stone  pine 

3 

0.235 

0.268 

0.105 

0.009 

(0.207-0.261) 

(0.264-0.275) 

(0.096-0.119) 

(0.004-0.014) 

Whitebark  pine 

1 

0.121 

0.130 

0.185 

0.256 

(0.111-0.129) 

(0.166-0.220) 

(0.253-0.261) 

'N  =  number  of  localities  studied. 


And  finally,  the  low  differentiation  between  stone  pine 
populations  can  be  considered  the  result  of  a  specific  mode 
of  seed  dispersal  by  nutcrackers  able  to  carry  their  seeds 
over  very  long  distances  (up  to  20  km,  according  to  some 
observations).  The  efficiency  of  this  mode  and  its  influence 
on  the  population  structure  were  repeatedly  emphasized 
by  researchers  studying  relationships  between  nutcrackers 
and  whitebark  pine,  Swiss  stone  pine,  limber  pine  {Pinus 
flexilis  James),  and  pinyons  (subsection  Cembroides) 
(Carsey  and  Tomback  1992;  Furnier  and  others  1987; 
Lanner  1980,  1990;  Schuster  and  Mitton  1991;  Tomback 
and  others  1990,  1992). 

Cluster  and  Principal  Coordinates  Analysis — 

Despite  the  low  values  of  genetic  distances,  UPGMA  cluster- 
ing of  stone  pine  samples  based  on  these  distances  reflects, 
in  general,  their  geographical  origin.  More  detailed  analy- 
ses have  been  performed  on  Siberian  stone  pine,  which  has 
been  studied  more  than  other  stone  pines.  We  can  tenta- 
tively divide  all  11  Siberian  stone  pine  localities  into  three 
regions,  according  to  their  ecology  and  geographic  location. 
The  first  of  them  can  be  called  "Southern  Siberia  Moimtains" 
and  includes  one  population  from  the  Altai  Mountains  (Ya) 
and  five  localities  in  the  western  Sayan  (FK,  M,  MK,  SR, 
and  L);  the  second  can  be  called  "Western  Siberia  Plain" 
and  includes  populations  from  Tomsk  region  (S,  Z-1,  and 
Z-3);  the  third,  "Northern  Siberia,"  includes  populations 
from  Khanty-Mansi  Autonomous  District  (NYa)  and  from 
the  north  of  Krasnoyarsk  Territory  (TU).  The  dendrogram 
obtained  corresponds  exactly  to  this  classification  (fig.  1). 
These  three  regions  form  three  different  genetic  groups 
(clusters)  of  populations. 

Korean  pine  samples  also  form  clusters  in  accordance 
with  their  population  localization  (two  geographically  more 
closely  spaced  stands,  MKh  and  Kho,  have  turned  out  to 
be  genetically  more  similar).  The  same  is  true  of  Japanese 
stone  pine  populations  (fig.  1).  Principal  coordinates  analy- 
ses based  on  the  matrix  of  Nei's  genetic  distances  between 
Siberian  stone  pine  populations  also  have  been  performed. 
This  multidimensional  analysis  has  produced  almost  the 
same  result  as  the  cluster  analysis.  Genetic  relationships 
between  Siberian  stone  pine  populations,  graphically  pre- 
sented in  two-dimensional  space  of  the  first  two  principal 
coordinates,  have  perfectly  reflected  the  real  geographical 


distribution  of  these  populations  (fig.  2).  Principal  coordi- 
nates and  cluster  analyses  complement  each  other  and  have 
produced  almost  the  same  result  in  our  case,  but  the  first, 
probably,  gives  more  information  about  population  genetic 
relationships  due  to  better  data  presentation.  Multivariate 
analysis  techniques  are  supposed  to  provide,  basically,  bet- 
ter possibility  for  revealing  genetic  differentiation  in  coni- 
fers, as  compared  to  univariate  methods  (Krutovskii  and 
Politov  1992;  Yeh  and  others  1985). 

Statistically  significant  positive  correlation  between  geo- 
graphical and  genetic  distances  has  been  found  for  Siberian 
stone  pine  populations  (r  =  0.844,  P  <  0.001).  Such  correla- 
tion was,  however,  absent  for  samples  from  closely  spaced 
localities  of  the  Malyi  Kebezh  River  Basin.  The  limited 
number  of  samples  studied  has  not  allowed  such  analysis 
for  other  stone  pine  species,  but  analogous  results  have 


0.010 


0.005 
I 


D 

F 

L 
M 
MK 

Southern 

Siberia 

Mountains 

YA 

SR  _ 

Z-1 

Z-3 

Western 

Siberia 

Plain 

NYA  _ 

Northern 
Siberia 

MKH~ 
KHO_ 

1  Khabarovsk" 
Territory 

SA 

Primorski!  - 
Territory 

GR 
UT 

K  - 

Kamchatka 
Peninsula 

Siberian 

stone  pine 


Korean 
stone  pine 

Japanese 
stone  pine 


Figure  1 — UPGMA  phenograms  of  stone  pine 
populations  based  on  Nei's  genetic  distances, 
D,  (Nei  1972).  (See  full  names  and  origins  of 
populations  in  table  1 .) 


23 


NYA 


Southern 
Siberia 
Mountains 


Western 
Siberia  Plain 


-0.15 


Figure  2 — Projections  of  Siberian  stone  pine 
populations  in  two-dimensional  space  of  the  first 
two  principal  coordinates,  X  and  Y,  according  to 
Nei's  genetic  distances,  D,  (Nei  1972).  (See  full 
names  and  origins  of  populations  in  table  1 .) 


also  been  obtained  for  some  other  pine  species  (Furnier 
and  Adams  1986;  Steinhoff  and  others  1983;  Wheeler  and 
Guries  1982),  including,  probably,  that  isolation  by  distance 
(Nei  1975),  introgressive  hybridization  (Millar  1983),  or 
gradiently  changing  selection  (Bergmann  1978)  could  play 
a  significant  role  in  genetic  differentiation  of  some  pine 
species. 

We  have  also  analyzed  correlation  between  geographical 
and  Nei's  distances  calculated  for  Siberian  stone  pine  popu- 
lations using  data  on  individual  loci  (allele  frequencies  of 
eight  loci  in  the  effective  pollen  pool  have  been  used).  For 
loci  Adh-1,  Dia-2,  Fe-2,  Pgm-1,  and  Skdh-1  correlation  was 
significant.  The  other  loci  (Lap-3,  Mdh-2,  and  Pgi-2)  dis- 
played lack  of  correlation.  This  may  suggest  a  selective  role 
of  certain  loci.  Patterns  of  spatial  differentiations  seem  to 
be  determined  by  complex  interaction  of  gene  flow  and  selec- 
tion. Distribution  of  genetic  variability  may  be  also  affected 
by  "evolutionary  footprints,"  such  as,  for  example,  Siberian 
population  descent  from  different  refugiums. 

INTERSPECIFIC  DIFFERENTIATION 

Genetic  Distances  Between  Stone  Pine  Species — 

We  have  estimated  both  genetic  distances,  D^j,  and  D,  be- 
tween all  stone  pine  species  based  on  16  isozyme  loci  data, 
but  only  D  values  are  presented  in  table  3,  because  of  space 
limits.  The  smallest  genetic  distances  have  been  found 
between  Siberian  and  Swiss  stone  pines  (D  =  0.052-0.077). 
They  are  just  a  little  larger  than  those  between  subspecies 
of  lodgepole  pine  (Pinus  contorta  Dougl.  ex  Loud.),  D  =  0.008- 
0.019,  approximately  0.012  on  the  average  (Wheeler  and 
others  1983),  similar  to  the  distances  between  subspecies 
of  brutian  pine  {Pinus  brutia  Ten.),  D  =  0.03-0.09  (Conkle 
and  others  1988),  and  a  little  smaller  than  distances  be- 
tween two  very  closely  related  species,  jack  pine  (Pinus 
banksiana  Lamb.)  and  lodgepole  pine,  which  have  a  wide 
zone  of  introgressive  hybridization,  D  =  0.097,  according  to 


the  study  by  Dancik  and  Yeh  (1983),  or  D  =  0.108,  according 
to  Wheeler  and  Guries  (1987).  Nei's  distances  estimated 
between  other  pairs  of  stone  pine  species  (D  =  0.121-0.268, 
table  3)  are  close  to  the  values  reported  for  pine  species 
of  subsections  Contortae:  0.078-0.158  (Wheeler  and  others 
1983)  and  Oocarpae:  0.11-0.27  (Millar  and  others  1988). 
Approximately  the  same  value,  D  =  0.27-0.43,  has  been 
reported  for  Pinus  brutia-Pinus  halepensis  comparison 
(Conkle  and  others  1988).  We  have  estimated  the  average 
genetic  distance  between  species  fi-om  the  same  subsections 
using  a  few  references  (Conkle  and  others  1988;  Millar  and 
others  1988;  and  some  others).  It  equaled  0.185,  on  the  av- 
erage, for  14  pine  species  and  surprisingly  corresponded  to 
the  average  distance  estimated  for  all  stone  pine  species  in 
our  study — D  =  0.180,  and  for  four  Eurasian  stone  pine  spe- 
cies studied  by  Shurkhal  and  others  (1991a,b)— D  =  0.138 
(for  comparison,  D  =  0.185,  on  the  average,  for  four  Eurasian 
stone  pines  used  in  our  study).  The  small  discrepancy  be- 
tween the  latter  two  estimates  is  not  surprising,  taking  into 
account  the  sampling  strategy  used  by  Shurkhal  and  others 
(1991a,b;  1992).  They  have  analyzed  only  a  few  trees  of 
each  species  (mostly,  from  Botanical  Garden  collections), 
which  should  inevitably  bias  genetic  distance  estimates  to 
smaller  values,  due  to  a  higher  probability  of  random  fixa- 
tion of  most  common,  frequent  alleles  in  small  samples  of 
trees  from  different  species. 

Levels  of  interspecific  genetic  differentiation  obtained 
between  stone  pines  strongly  confirm  their  taxonomical 
status  as  closely  related  species  of  the  same  subsection. 
Nei's  genetic  distances  obtained  for  species  belonging  to 
different  subsections  or  sections  are  usually  higher  than 
0.3  and  vary  between  0.3  and  0.8  (Shurkhal  and  others 
1991a,b;  1992).  According  to  phylogenetic  analysis  carried 
out  by  Shiu-khal  and  others  (1991a,b;  1992)  using  isoz)Tne 
loci,  fotir  stone  pine  species  of  subsection  Cembrae  (white- 
bark  pine  has  not  been  analyzed)  and  three  white  pine  spe- 
cies of  subsection  Strobi,  Balkan  or  Macedonian  pine  (Pinus 
peuce  Griseb.),  blue  or  Himalayan  pine  (Pinus  griffithii 
McCleland),  and  eastern  white  or  We3maouth  pine  (Pinus 
strobus  L.),  form  separate  clusters.  The  Nei's  genetic  dis- 
tance between  these  two  very  closely  related  subsections 
equals  0.278,  on  the  average  (calculated  by  us,  using  the 
data  in  table  5  from  Shurkhal  and  others  1991b).  This  is 
almost  two  times  larger  than  that  between  Eurasian  stone 
pines  (D  =  0.138). 

The  lack  of  species-specific  diagnostic  loci  among  stone 
pines  also  confirms  the  correctness  of  uniting  them  in  the 
same  separate  subsection  Cembrae.  Only  Siberian  and 
Korean  stone  pines,  on  locus  Gdh,  and  Swiss  and  Korean 
stone  pines,  on  locus  Mdh-4,  do  not  have  common  alleles. 
Besides,  Korean  stone  pines  have  a  unique  electrophoretic 
pattern  of  LAP  enzyme  indicating  the  existence  (or  expres- 
sion) of  an  additional  Lap-4  locus. 

Loci  Adh-1,  Gdh,  Got-3,  Mdh-4,  and  Pgi-2  make  a  most 
significant  contribution  to  interspecific  differentiation  of 
stone  pines,  but  totally  only  39  percent  of  the  whole  sub- 
section Cembrae  genetic  variation  belongs  to  interspecific 
differences  (table  2). 

Thus,  our  data  give  a  strong  genetic  evidence  of  common, 
apparently  monophyletic  stone  pines  origin  and  compara- 
tively recent  time  of  their  divergence.  Subsection  Cembrae 
seems  to  be  a  valid  taxon. 


24 


Phylogenetic  Relationships  Inferred  From  Genetic 
Distances — Many  publications  have  been  devoted  to  stud- 
ies of  taxonomic  and  phylogenetic  relationships  between 
stone  pines  using  morphological,  physiological,  ecological, 
and  other  traits.  Our  study  was  the  first  one  to  use  genetic 
markers,  isozyme  loci,  for  analysis  of  phylogenetic  relation- 
ships between  all  five  stone  pine  species.  These  species  are 
traditionally  united  in  subsection  Cembrae,  according  to 
some  characteristic  features  (Critchfield  and  Little  1966; 
Farjon  1984;  Little  and  Critchfield  1969;  Shaw  1914).  The 
overwhelming  majority  of  researchers  consider  the  main  spe- 
cific features  of  this  subsection,  indehiscent  cones  and  wing- 
less seeds,  to  be  inherited  fi*om  a  common  ancestor.  These 
traits  seem  to  be  a  relatively  recent  acquisition.  Their  ap- 
pearance is  considered  to  be  a  result  of  coevolution  of  stone 
pines  and  nutcrackers  of  genus  Nucifraga  foraging  their 
seeds  and  playing  the  main  role  in  their  maintenance  and 
dispersal  (Lanner  1980,  1982,  1990;  Tomback  1983; 
Tomback  and  Linhart  1990;  Tomback  and  others  1990). 

Thus,  the  subsection  Cembrae  is  regarded  as  a  natural 
monophyletic  taxon.  Nevertheless,  there  also  exists  a  dif- 
ferent point  of  view  (Critchfield  1986),  according  to  which 
traits  shared  by  all  stone  pines  are  insufficiently  confirmed 
as  homologous.  Data  on  artificial  hybridization  appear  to 
be  rather  contradictory  (Critchfield  1986).  The  subsection 
is  believed  to  have  polyphyletic  origin  and,  therefore,  is 
not  a  valid  taxon  (see  for  discussion,  Lanner  1990).  Millar 
and  Kinloch  (1991)  also  have  pointed  out  a  nimiber  of  im- 
solved  problems  in  taxonomy  and  phylogeny  of  subsection 
Cembrae. 

Molecular  and  biochemical  genetic  markers  have  proven 
to  be  a  useful  tool  for  study  of  biosystematic  and  phyloge- 
netic relationships  in  conifers  (Strauss  and  others  1992). 
We  believe  that  application  of  biochemical  genetic  markers 
like  isozymes  can  help  shed  light  on  the  evolution  of  stone 
pine  species.  The  dendrograms  produced  by  NJM  and 
UPGMA  clustering  using  D^h  and  D  matrix  (table  3)  based 
on  16  isozyme  loci  have  apparently  revealed  real  phyloge- 
netic relationships  between  five  stone  pine  species  (fig.  3). 

According  to  this  dendrogram,  Siberian  and  Swiss  stone 
pines  appear  to  be  a  most  closely  related  pair  of  species 
(D  =  0.065).  This  was  not  imexpected,  because  of  the  well- 
known  surprisingly  high  morphological  similarity  (Bobrov 
1978)  and  successful  artificial  hybridization  (Critchfield 
1986)  between  them.  Practically  all  traditional  botanical 
classifications  note  the  absence  of  diagnostic  morphological 
traits,  which  means  that  interspecific  differences  are  more 
quantitative  than  qualitative.  According  to  paleobotanical 
(mainly  palynological)  data,  Bobrov  (1978)  considered  these 
two  species  as  two  now  separately  located  marginal  parts 
of  a  former  single  ancestral  species  spread  on  a  vast  range 
area,  from  the  western  Alps  to  the  northeastern  Yakutia 
(eastern  Siberia),  whose  pollen  has  been  foimd  in  Paleocene, 
Pleistocene,  and  later  deposits. 

As  Bobrov  (1978)  indicates,  "Siberian  stone  pine  must  be 
regarded  as  the  initial,  more  ancient  species,  formerly  more 
widely  distributed,  whereas  the  West  European  stone  pine, 
Pinus  cembra,  is  evolutionarily  still  very  yoimg  and  has  be- 
come geographically  isolated  in  the  recent  past,  being,  there- 
fore, hardly  disting\iishable  from  its  Siberian  ancestors." 

Breakup  of  this  ancient  range,  its  division  into  two  parts, 
and  decrease  of  its  western  part  occurred,  most  likely,  as 


UPGMA 


Isozyme  loci 


L 


JL 


-L 


225    200    175   150   125    100    75     50  25 


UPGMA 


cpDNA  RFLP 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 

D  (X10-3) 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 


J. 


± 


± 


J. 


90    80     70    60     50    40    30     20  10 


UPGMA 


isozyme  loci 


C3 


d  (X10-5) 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 


_L 


_L 


_L 


I 


405    360  315    270  225    180   135    90  45 


Dch(x10-3) 


NJM  isozyme  loci 

Siberian  stone  pine 


C Siberian 


± 


± 


Swiss  stone  pine 

Whitebari<  pine 

— —  Korean  stone  pine 
-■i-v— i— i^^—  Japanese  stone  pine 

-J  \  \  lD,,(x10-3) 


45    90     135  180    225    270  315  360  405 


NJM 


Swiss  stone  pine 
Siberian  stone  pine 

Korean  stone  pine 


cpDNA  RFLP 

Whitebark  pine 


I  L 


J  L 


15 


30 


45 


65 


Japanese  stone  pine 
-Jd  (X10-5) 


Figure  3 — UPGMA  and  NJM  dendrograms  of  five 
stone  pine  species  (subsection  Cembrae)  based  on 
Cavalli-Sforza  and  Edwards'  (1967)  and  Nei's  (Nei 
1972)  genetic  distances,     and  D,  calculated  from 
isozyme  data,  and  on  d  (Nei  1 987b;  Nei  and  Miller 
1990),  calculated  from  chloroplast  (cp)  DNA  restric- 
tion fragment  length  polymorphism  (RFLP)  data 
(Krutovskii  and  Wagner  1993). 


25 


late  as  in  the  Holocene  and  was  associated  with  its  substi- 
tution in  the  central  part  of  the  area  by  other  species  due 
to  the  global  warming.  The  time  of  divergence  (10,000  to 
25,000  years),  estimated  from  the  genetic  distance  between 
Siberian  and  Swiss  stone  pines  corresponds,  approximately, 
to  these  events  (for  more  details  see  Krutovskii  and  others 
1990).  Thus,  the  low  level  of  divergence  between  Siberian 
and  Swiss  stone  pines  can  be  explained  by  the  short  time 
of  isolation. 

Nevertheless,  a  considerable  amount  of  genetic  difference 
has  accumulated  during  this  time — genetic  distances  be- 
tween Siberian  and  Swiss  stone  pines  are  approximately 
10  times  larger  than  those  between  populations  of  the  same 
species  (table  3)  and  are  comparable  with  distances  reported 
by  Conkle  and  others  (1988)  for  subspecies  and  very  closely 
related  pine  species.  Our  data  have  also  been  recently  con- 
firmed by  Goncharenko  and  others  (1992)  and  Shurkhal  and 
others  (1991a,b;  1992).  Their  estimates  of  Nei's  genetic  dis- 
tance between  Siberian  and  Swiss  stone  pines  equaled 
0.030  and  0.042,  correspondingly.  Thus,  Swiss  stone  pine 
seems  to  be  a  highly  diverged  geographical  race  or  subspe- 
cies of  Siberian  stone  pine,  or  a  species  in  statu  nascendi. 

Unexpectedly  relatively  small  differentiation  found  be- 
tween Japanese  and  Korean  stone  pines  (D  =  0.105)  makes 
us  search  for  a  new  approach  to  understanding  their  origin. 
These  species,  in  fact,  have  little  in  common  in  morphology 
of  both  vegetative  and  generative  organs,  but  according  to 
the  study  of  parenchjnnal  tissue  of  needle  cell  walls,  stone 
pine  species  and  some  closely  related  species  can  be  divided 
into  two  t3T)es:  Siberian  and  Swiss  stone  pine  form  the  first 
type,  while  Japanese  and  Korean  stone  pines  belong  to  the 
second  type,  but  to  different  subtypes  (Litvintseva  1974). 
Perhaps,  the  evolutionary  pathways  of  these  species  have 
truly  separated  from  a  common  ancestor  relatively  recently. 
Another  explanation  of  their  genetic  similarity  is  gene  ex- 
change through  introgressive  hybridization,  which  could 
have  occurred  (or  seldom  occurs),  taking  into  account  the 
present  sympatry  of  these  species  and,  as  a  rule,  poorly 
developed  mechanisms  of  reproduction  isolation  between 
closely  related  pine  species.  More  detailed  discussion  can 
be  foimd  in  Krutovskii  and  others  (1990),  but  recently  ob- 
tained data  also  confirm  the  division  of  Eurasian  stone  pines 
by  isozyme  loci  analysis  into  two  groups:  Siberian  and  Swiss 
stone  pines  (D  =  0.030-0.042),  on  one  hand,  and  Korean  and 
Japanese  (D  =  0.089-0.143),  on  the  other  hand  (Goncharenko 
and  others  1992;  Shurkhal  and  others  1991a,b;  1992). 

A  most  interesting  problem,  from  our  point  of  view,  is  ex- 
planation of  the  whitebark  pine's  closer  affinity  to  Siberian 
and  Swiss  stone  pines  (D  =  0.121-0.130)  than  to  Japanese 
stone  pine  (D  =  0.185),  although  quite  opposite  relation- 
ships could  be  suspected  considering  these  species'  present 
area.  Among  stone  pines  Siberian  and  Swiss  stone  pines 
are,  probably,  genetically  most  similar  to  the  original  an- 
cient stone  pine,  due  to  large  refugiums  in  the  Alps  and 
Siberian  (Altai  and  Sayan)  Mountains,  which  could  help 
them  save  a  great  deal  of  ancient  genetic  variation;  and 
whitebark  pine  also  could  keep  more  genetic  similarity  with 
the  ancient  species  than  Korean  and  Japanese  stone  pines. 
Another  explanation  could  be  that  stabilizing  selection  sup- 
ports genetic  similarity  between  whitebark  pine  and  Siberian 
and  Swiss  stone  pines,  because  whitebark  pine  ecologically 


has,  probably,  more  in  common  with  them  than  with  Korean 
and  Japanese  pines. 

However,  the  dendrogram  obtained  by  NJM,  and  based 
on  Cavalli-Sforza  and  Edwards'  (1967)  distance,  D^^,  has 
revealed  a  kind  of  "intermediate"  position  of  whitebark  pine 
between  two  stone  pine  groups,  Swiss  and  Siberian  stone 
pines,  on  one  hand,  and  Korean  and  Japanese  stone  pines, 
on  the  other  (fig.  3).  In  these  cases  NJM  gives  more  an  in- 
tuitively realistic  phylogenetic  tree.  It  is  important  to  note 
that  both  UPGMA  dendrograms  based  on  D^,,  and  Nei's  dis- 
tance, D,  have  practically  the  same  topology,  but  they  differ 
from  the  NJM  dendrogram.  Thus,  one  can  conclude  that 
assumption  of  equal  evolutionary  rates  of  stone  pine  spe- 
cies can  be  erroneous. 

Millar  and  Kinloch  (1991)  discuss  several  theoretically 
possible  variants  of  whitebark  pine  origin:  (1)  from  ancestral 
Eurasian  stone  pine  with  penetration  into  North  America 
during  the  Mesozoic  (before  Laurasia  segregation).  Ter- 
tiary, or  Quartemary  (through  the  Bering  land  bridge)  time; 
(2)  fix)m  one  of  the  American  pines  of  subsection  Strobi.  Now 
we  can  finally  solve  this  problem.  Nei's  genetic  distances 
(D)  between  whitebark  pine  and  Eurasian  stone  pines  equal 
approximately  0.121-0.256  (table  3).  The  time  of  divergence 
it)  between  whitebark  pine  and  Eurasian  stone  pines,  theo- 
retically estimated  from  these  values,  is  0.6-1.3  million  years, 
using  calibration  coefficient  (k)  5x10^  in  the  equation  t  =  kD 
(Nei  1975;  1987a,b).  This  corresponds  quite  well  to  the  ap- 
proximate geological  time  of  the  Bering  Strait  opening — 
1.8-3.5  million  years  ago — estimated  by  different  methods 
(see  references  in  Grant  1987).  Consequently,  this  event, 
the  Bering  Strait  opening,  can  be,  obviously,  considered 
as  the  beginning  of  genetic  differentiation  between  North 
American  and  Eurasian  pines  due  to  geographical  isolation. 

Thus,  the  hypothesis  of  a  relatively  recent  penetration  of 
ancient  stone  pines  into  North  America  through  the  Bering 
lemd  bridge  in  the  Pliocene  seems  most  preferable.  If  we 
assumed  the  hypothesis  that  a  more  ancient  penetration 
of  stone  pines  through  a  "Euramerican"  connection  had 
occurred  during  Mesozoic  time  (Millar  and  Kinloch  1991), 
much  more  differentiation  between  whitebark  pine  and 
Eurasian  stone  pines  should  have  been  found  in  our  study. 
The  assumption  of  whitebark  pine  origin  from  Strobi  pines 
must  be  rejected,  according  to  our  data,  but  phylogenetic 
relationships  of  stone  pines  with  other  relative  species  from 
subsections  Strobi  and  Cembroides  certainly  need  much 
more  analysis.  The  first  steps  in  this  direction  have  been 
made  recently  using  cpDNA  markers. 

Chloroplast  DNA  Data  and  Pine  Phylogeny— Only 
a  few  studies  on  pine  phylogeny  using  DNA  markers  have 
been  made.  This  allows  us  to  draw  a  short  comparison  of 
the  Millar  and  Kinloch  (1991)  results  with  our  data,  pre- 
sented here,  paying  special  attention  to  phylogenetic  rela- 
tionships of  stone  pines  with  some  other  closely  related 
species. 

Chloroplast  DNA  (cpDNA)  restriction  fragment  length 
polymorphism  (RFLP)  of  the  five  stone  pine  species  and  four 
white  pine  species  of  a  closely  related  subsection  Strobi — 
limber  pine,  sugar  pine  (P.  lambertiana  Dougl.),  western 
white  pine  (P.  monticola  Dougl.),  and  eastern  white  pine — 
have  been  recently  studied  by  Krutovskii  and  Wagner  (1993). 
Data  have  been  obtained  by  molecular  hybridization  of  27 


26 


probes,  representing  the  entire  chloroplast  genome  of 
P.  contorta,  with  restriction  fragments  from  20  endonu- 
cleases.  The  total  nimiber  of  restriction  fragments  per 
sample  varied  from  378  to  422. 

Chloroplast  DNA  variation  among  trees  within  species 
was  not  pronounced  (D  =  0.0-0.0013)  and  was  comparable 
with  variation  between  populations  within  species,  but 
cpDNA  diversity,  in  general,  was  quite  high — 50  to  100 
percent  individual  cpDNA  belonged  to  different  haplotypes 
(for  more  details  see  Krutovskii  and  Wagner  1993).  The 
dendrogram  constructed  on  Nei's  genetic  distances  (D), 
estimated  from  these  data  (Nei  1987b;  Nei  and  Miller 
1990),  has  shown  almost  the  same  phylogenetic  relation- 
ships between  stone  pine  species  as  established  on  isozyme 
loci  data  (fig.  3).  The  stone  and  white  pine  species  have 
been  imdoubtedly  divided  into  two  separate  groups  (fig.  4). 
It  is  interesting  to  note  that  NJM  dendrograms  give  better 
correspondence  between  cpDNA  and  isozyme  data.  Com- 
paring NJM  and  UPGMA  dendrograms  we  can  suppose 
that  assumption  of  homogeneity  of  evolutionary  rates  is 
less  proper  for  cpDNA  than  for  nuclear  (isozyme)  genes. 

The  cpDNA  data  also  indicate  that  genetic  differentia- 
tion is  more  developed  among  the  four  Strobi  species  than 


among  the  Cembrae  species.  The  same  concliision  has  been 
made  by  Shurkhal  and  others  1991a,b;  1992)  on  the  isozyme 
data. 

Chloroplast  DNA's  of  whitebark  pine  and  Eurasian  stone 
pine  species  were  remarkably  similar.  This  similarity,  to- 
gether with  clear  cpDNA  divergence  between  the  two  sub- 
sections, confirms  systematic  classification  of  whitebark 
pine  in  subsection  Cembrae,  and  not  in  subsection  Strobi. 
Phylogenetic  relationships  of  whitebark  pine  with  Siberian 
and  Swiss  stone  pines  revealed  by  cpDNA  RFLP  analysis 
are  surprisingly  consistent  with  the  conclusions  from  the 
analysis  of  16  isozyme  loci  (fig.  3).  Thus,  ovir  observation 
is  not  an  artifact,  because  two  independent  sets  of  genetic 
markers  (isozymes  and  cpDNA  RFLPs)  have  produced  the 
same  conclusions.  Moreover,  Szmidt  (1991)  notes  that  the 
present  geographic  distribution  of  some  forest  tree  species 
often  shows  little  correlation  with  the  degree  of  cpDNA  dif- 
ferentiation among  them,  because  the  present  geographic 
distribution  of  some  species  is  of  relatively  recent  origin. 

The  data  obtained  by  Szmidt  and  others  (1988),  who  used 
four  restriction  enzymes  for  cpDNA  RFLP  analysis  of  12 
pine  species,  including  two  stone  pine  species  (Swiss  stone 
and  whitebark  pines)  and  two  white  pine  species  (western 


UPGMA 


cpDNA  RFLP 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 
Western  white  pine 
Limber  pine 
Sugar  pine 
Eastern  white  pine 

d  (xlO-3) 


Stone 
Pines 


White 
Pines 


30 


20 


10 


NJM 


J  Sit 

it 


Siberian  stone  pine 
Swiss  stone  pine 
Whitebark  pine 
Korean  stone  pine 
apanese  stone  pine 

Western  white  pine 
Limber  pine 
Sugar  pine 


Stone 
Pines 


Eastern  white  pine 
I  id(x10-3) 


White 
Pines 


10 


20 


30 


Figure  4 — UPGMA  and  NJM  dendrograms  of  five  stone  pine  spe- 
cies (subsection  Cembrae)  and  four  white  pine  species  closely  re- 
lated to  them:  Pious  flexilis,  P.  lambertiana,  P.  monticola,  and  P. 
strobus  (subsection  Strobi),  based  on  Nei's  genetic  distances,  d, 
(Nei  1987b;  Nei  and  Miller  1990),  and  calculated  from  chloroplast 
(cp)  DNA  restriction  fragment  length  polymorphism  (RFLP)  data 
(Krutovskii  and  Wagner  1993). 


27 


and  eastern  white  pines),  also  confirm  our  results.  They 
evaluated  phylogenetic  relationships  between  these  species 
using  UPGMA  clustering  based  on  dissimilarity  index,  anal- 
ogous to  Nei's  D.  According  to  the  dendrogram  given  by 
Szmidt  £ind  others  (1988),  Swiss  stone  and  whitebgirk  pines 
form  a  very  tight  separate  cluster.  Western  white  pine  is 
distantly  linked  to  this  cluster,  which  exactly  corresponds 
to  our  data  (fig.  4).  Eastern  white  pine  forms  a  separate 
branch  but  is  closely  linked  to  the  group  of  Swiss  stone, 
whitebark,  and  western  white  pines. 

Analogous  data  have  been  obtained  by  Strauss  and 
Doerksen  (1990)  who  studied  one  sample  fi"om  each  of  18 
to  19  pine  species  representing  14  to  15  subsections  using 
eight  restriction  enzjnnes  and  17  cloned  DNA  fragments 
(mostly  fi"om  Pseudotsuga  menziesii).  At  least  65  percent 
of  the  chloroplast  genome  was  covered  by  the  combination 
of  the  probes  used.  The  samples  have  also  included  two 
stone  pine  species  (Korean  stone  and  whitebark  pines)  and 
two  white  pine  species  (blue  and  sugar  pines).  Distances 
between  the  samples  were  estimated  via  shared  characters. 
The  stone  £uid  white  pine  species  were  extremely  similar 
to  one  another,  with  distances  ranging  fi^om  0.0  to  0.018. 
They  form  one  cluster.  The  authors  have  mentioned  (data 
have  not  been  presented)  that  other,  more  limited,  restric- 
tion fi'agment  studies  of  section  Strobus,  including  three 
stone  (Swiss,  Siberian,  and  Japanese)  and  three  white  pines 
(Armand's  [P.  armandii  Franch.],  limber,  and  western  white), 
have  given  the  same  indication  of  high  uniformity.  Unfor- 
tunately, Strauss  and  Doerksen  (1990)  have  not  been  able 
to  distinguish  stone  and  white  pines,  probably  because  a 
limited  number  of  restriction  enzymes  or  DNA  probes  have 
been  used  in  their  research. 

Certainly,  stone  pines  and  other  closely  related  species 
need  more  genetic  analysis  using  both  isozyme  and  DNA 
markers.  Many  fascinating  aspects  of  their  evolution  are 
sure  to  remain  hidden,  and  many  discoveries  are  still  w£dt- 
ing  for  research. 

CONCLUSIONS 

The  main  conclusions  based  on  our  data  are: 

1.  Both  biochemical  (isozymes)  and  molecular  (cpDNA) 
genetic  markers  are  very  useful  and  informative  for  study 
of  population  genetic  structure  and  phylogeny  of  stone  pine 
and  other  closely  related  species,  and  they  complement  each 
other. 

2.  In  spite  of  the  high  level  of  isozyme  loci  and  cpDNA 
diversity  among  stone  pine  species,  genetic  differentiation 
between  the  populations  is  not  pronounced,  but  specific. 
As  a  minimum,  the  genetic  differentiation  reflects  their 
geographical  and  ecological  peculiarities.  Thus,  it  is  neces- 
sary to  study  populations  fi'om  the  whole  of  the  stone  pine 
species  area  for  careful  description  of  gene  distribution, 
taking  into  account  even  very  small  genetic  differences 
between  populations  for  developing  gene  conservation  and 
reforestation  programs. 

3.  According  to  both  isozyme  and  cpDNA  restriction  fi'ag- 
ment  analyses,  P.  albicaulis  undoubtedly  belongs  to  Cembrae 
species  and  not  to  Strobi  species. 

4.  Stone  pine  species  consist  of  a  compact  group  of  very 
closely  related  and  genetically  similar  species,  obviously 


having  monophyletic  origin,  supposedly,  fi*om  ancient 
Siberian  stone  pine.  Thus,  subsection  Cembrae  should 
be  considered  as  a  valid  taxon. 

ACKNOWLEDGMENTS 

We  are  very  gratefiil  to  V.  N.  Vorobjev  and  N.  A.  Vorobjeva 
(Institute  of  Ecology  of  Natural  Complexes,  Tomsk,  Russia), 
L.  I.  Milyutin  and  his  laboratory  staff  (Institute  of  Forest 
and  Timber,  Krasnoyarsk,  Russia),  G.  R.  Furnier  (Univer- 
sity of  Minnesota,  Minneapolis,  MN,  U.S.A.),  R,  T.  Gut  (For- 
est Technology  Institute,  Lvov,  Ukraine),  and  P.  Razumov 
(Far  East  Forest  Institute,  Khabarovsk,  Russia)  for  their 
help  in  collecting  and  kindly  providing  Cembrae  species  seed 
samples  for  isozjnne  analysis.  We  are  th£inkful  to  Yu.  S. 
Belokon'  (Institute  of  General  Genetics,  Moscow,  Russia) 
for  very  helpfiil  assistance  in  electrophoretic  ansdysis,  R.  M. 
Lanner  (Utah  State  University,  Logan,  UT,  U.S.A.),  D.  F. 
Tomback  (University  of  Colorado  at  Denver,  Denver,  CO, 
U.S.A.),  and  C.  Millar  (Institute  of  Forest  Grenetics,  Berkeley, 
CA,  U.S.A.)  for  very  useful  comments  on  pine  phylogeny. 
We  also  appreciate  H.-R.  Gregorius'  invaluable  comments 
that  helped  to  substantially  improve  the  manuscript. 

REFERENCES 

Altukhov,  Y.  P.  1990.  Population  genetics:  diversity  and  sta- 
bility. New  York:  Harwood  Academic  Publishers.  352  p. 

Altukhov,  Y.  P.;  Krutovskii,  K  V.;  Dukharev,  V.  A.; 
Larionova,  A.  Y.;  Politov,  D.  V.;  Ryabokon',  S.  M.  1989. 
Biochemicgd  population  genetics  of  woody  forest  species. 
In:  Iroshnikov,  A.  I.,  ed.  International  symposium  on  for- 
est genetics,  breeding  and  physiology  of  woody  plants: 
proceedings;  1989  September  24-30;  Voronezh.  Moscow: 
The  State  Forest  Committee  of  the  USSR:  21-29. 

Bergmginn,  F.  1978.  The  allelic  distribution  of  an  acid  phos- 
phatase locus  in  Norway  spruce  {Picea  abies)  £dong  simi- 
lar climatic  gradients.  Theoretical  and  Applied  Genetics. 
24:  57-64. 

Bobrov,  E.  G.  1978.  Forest-forming  conifers  of  the  USSR. 
Moscow:  Nauka.  190  p.  [In  Russian]. 

Carsey,  K.  S.;  Tomback,  D.  F.  1992.  Growth  form  distribu- 
tion and  genetic  relationships  in  tree  clusters  of  Pinus 
flexilis,  a  bird-dispersed  pine.  Oecologia.  [In  press.] 

Cavalli-Sforza,  L.  L.;  Edwards,  A.  W.  F.  1967.  Phylogenetic 
analysis:  models  and  estimation  procedures.  American 
Journal  of  Human  Genetics.  19(2):  233-257. 

Conkle,  M.  T.;  Schiller,  G.;  Grunwald,  C.  1988.  Electro- 
phoretic analysis  of  diversity  and  phylogeny  of  Pinus 
brutia  £ind  closely  related  taxa.  Systematic  Botany. 
13(3):  411-424. 

Crawford,  D.  J.  1983.  Phylogenetic  and  systematic  infer- 
ences fi'om  electrophoretic  studies.  In:  Tanksley,  S.  D.; 
Orton,  T.  J.,  eds.  Isoz)nnes  in  plant  genetics  and  breed- 
ing. Part  A.  Amsterdam:  Elsevier  Scientific  Publishers: 
257-287. 

Critchfield,  W.  B.  1986.  Hybridization  and  classification 
of  the  white  pines  (Pinus  section  Strobus).  Taxon.  35(4): 
647-656. 

Critchfield,  W.  B.;  Little,  E.  L.  1966.  Geographic  distribu- 
tion of  the  pines  of  the  world.  Washington,  DC:  U.S.  De- 
partment of  Agriculture.  99  p. 


28 


Dancik,  B.  P.;  Yeh,  F.  C.  1983.  Allozyme  variability  and 
evolution  of  lodgepole  pine  {Pinus  contorta  var.  latifolia) 
and  jack  pine  (Pinus  banksiana)  in  Alberta.  Canadian 
Journal  of  Genetics  and  Cytology.  25:  57-64. 

Farjon,  A.  1984.  Pines:  drawings  and  descriptions  of  the 
genus  Pinus.  Leiden:  E.  J.  Brill.  204  p. 

Farris,  J.  S.  1972.  Estimating  phylogenetic  trees  from  dis- 
tance matrices.  American  Naturalist.  106(951):  645-668. 

Fumier,  G.  R.;  Adams,  W.  T.  1986.  Geographic  patterns  of 
allozyme  variation  in  Jeffrey  pine.  American  Journal  of 
Botany.  73(7):  1009-1015. 

Fumier,  G.  R.;  Knowles,  P.;  Clyde,  M.  A.;  Dancik,  B.  P.  1987. 
Effect  of  avian  seed  dispersal  on  the  genetic  structure  of 
whitebark  pine  populations.  Evolution.  41(3):  607-612. 

Goncharenko,  G.  G.;  Padutov,  V.  E.;  Silin,  A.  E.  1992.  Popu- 
lation structure,  gene  diversity,  and  differentiation  in  natu- 
ral populations  of  cedar  pines  (Pinus  subsection  Cembrae, 
Pinaceae)  in  the  USSR.  Plant  Systematics  and  Evolu- 
tion. 182(3-4):  121-134. 

(rottlieb,  L.  D.  1977.  Electrophoretic  evidence  and  plant 
systematics.  Annals  of  Missouri  Botanical  Garden.  64: 
161-180. 

Grant,  W.  S.  1987.  Genetic  divergence  between  congeneric 
Atlantic  and  Pacific  Ocean  fishes.  In:  Ryman,  N.;  Utter, 
F.,  eds.  Population  genetics  and  fishery  management. 
Seattle,  WA:  University  of  Washington  Press:  225-246. 

Hamrick,  J.  L.  1983.  The  distribution  of  genetic  variation 
within  and  among  natural  plant  populations,  hi:  Chambers, 
S.;  Schonewald-Cox,  C,  eds.  Genetics  and  wild  population 
management.  Reading,  PA:  Addison- Wesley:  335-349; 
501-509. 

Hamrick,  J.  L.;  Godt,  M.  J.  1989.  Allozyme  diversity  in  plant 
species.  In:  Brown,  A.  H.  D.;  Clegg,  M.  T.;  Kahler,  A.  L.; 
Weir,  B.  S.,  eds.  Plant  population  genetics  breeding  and 
genetic  resources.  Sanderland,  MA:  Sinauer  Associates: 
43-63. 

Hamrick,  J.  L.;  Loveless,  M.  D.  1986.  The  influence  of  seed 
dispersal  mechanisms  on  the  genetic  structure  of  plant 
populations.  In:  Estrada,  J.;  Fleming,  T.  H.,  eds.  Frugi- 
vores  and  seed  dispersal.  Dordrecht,  The  Netherlands: 
Dr.  W.  Junk  Publishers:  211-223. 

Hamrick,  J.  L.;  Mitton,  J.  B.;  Linhart,  G.  B.  1981.  Levels  of 
genetic  variation  in  trees:  influence  of  life  history  charac- 
teristics. In:  Conkle,  M.  T.,  ed.  Symposium  on  isozymes 
in  North  American  forest  trees  and  forest  insects:  proceed- 
ings. Berkeley,  CA:  U.S.  Department  of  Agricultiu-e,  For- 
est Service:  35-41. 

Krutovskii,  K.  V.;  PoHtov,  D.  V.;  Altukhov,  Yu.  P.  1987. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  I.  Genetic  control  of  isozyme  systems.  Genetika. 
23(12):  2216-2228.  [In  Russian].  [English  translation  pub- 
lished by  Plenvun  Publishing  Corp.  in  Soviet  Genetics; 
1988  June:  1568-1576]. 

Krutovskii,  K.  V.;  Pohtov,  D.  V.;  Altukhov,  Yu.  P.  1988. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  II.  Level  of  allozyme  variability  in  western 
Sayan.  Genetika.  24(1):  118-124.  [In  Russian].  [English 
translation  published  by  Plenum  Publishing  Corp.  in 
Soviet  Genetics;  1988  July:  81-86]. 

Krutovskii,  K  V.;  Pohtov,  D.  V.;  Altukhov,  Yu.  P.;  Milyutin, 
L.  I.;  Kuznetsova,  G.  V.;  Iroshnikov,  A.  I.;  Vorobjev,  V.  N.; 
Vorobjeva,  N.  A.  1989.  Genetic  variability  in  Siberian 
stone  pine,  Pinus  sibirica  Du  Tour.  IV.  Genetic  diversity 


and  differentiation  between  poptilations.  Genetika.  25(11): 
2009-2032.  [In  Russian].  [EngHsh  translation  published 
by  Plenum  Publishing  Corp.  in  Soviet  Genetics;  1990 
May:  1343-1362]. 

Krutovskii,  K.  V.;  PoHtov,  D.  V.;  Altukhov,  Yu.  P.  1990.  In- 
terspecific genetic  differentiation  of  Eurasian  stone  pines 
for  isoenzyme  loci.  Genetika.  26(4):  694-707.  [In  Russian]. 
[English  translation  published  by  Plenum  Publishing 
Corp.  in  Soviet  Genetics;  1990  October:  440-450]. 

Krutovskii,  K.  V.;  Politov,  D.  V.  1992.  Study  of  intra-  and 
interspecific  genetic  differentiation  between  E\u*asian 
cedar  pines  using  isozyme  loci  and  multidimensional 
analysis.  In:  Symposium  on  molecular  mechanisms  of 
genetic  processes:  proceedings.  Moscow:  Nauka  Publish- 
ers: 87-96.  [In  Russian]. 

Krutovskii,  K.  V.;  Wagner,  D.  B.  1993.  [Unpubhshed.]  Chlo- 
roplast  DNA  variation  and  phylogeny  in  Pinus  subsec- 
tions Cembrae  and  Strobi  (subgenus  Strobus).  Lexington, 
KY:  University  of  Kentucky,  Department  of  Forestry. 
[Working  draft]. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in  the 
ecology  and  evolution  of  limber  and  whitebark  pines.  In: 
6th  North  American  Forest  Biology  Workshop:  proceed- 
ings; 1980  August  11-13;  Edmonton,  AB.  Edmonton,  AB: 
University  of  Alberta:  15-47. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for  seed 
dispersal  by  Clark's  nutcracker.  Canadian  Journal  of 
Forest  Research.  12:  391-402. 

Lanner,  R.  M.  1990.  Biology,  taxonomy,  evolution,  and  geog- 
raphy of  stone  pines  of  the  world.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — sjonposium 
on  whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  14-24. 

Litvintseva,  M.  V.  1974.  Characteristics  of  needle  paren- 
chymal cell  structure  in  species  of  the  group  Cembrae 
of  the  genus  Pinus.  Botanicheski  Zhurnal.  [Leningrad]. 
59(10):  1501-1505.  [In  Russian]. 

Little,  E.  L.;  Critchfield,  W.  B.  1969.  Subdivisions  of  the 
genus  Pinas  (pines).  Res.  Rep.  1144.  Washington,  DC: 
U.S.  Department  of  Agriculture,  Forest  Service:  1-51. 

Loveless,  M.  D.;  Hamrick,  J.  L.  1984.  Ecological  determi- 
nants of  genetic  structure  in  plant  populations.  Annual 
Review  of  Ecology  and  Systematics.  15:  65-95. 

Millar,  C.  I.  1983.  A  steep  cline  in  Pinus  muricata.  Evolu- 
tion. 32(2):  311-319. 

Millar,  C.  I.;  Kinloch,  B.  B.  1991.  Taxonomy,  phylogeny,  and 
coevolution  of  pines  and  their  stem  rusts.  In:  Hiratsuka,  Y.; 
Samoil,  J.  K.;  Blenis,  P.  V.;  Crane,  P.  E.;  Laishley,  B.  L., 
eds.  Rusts  of  pine.  Inf.  Rep.  NOR-X-317.  Edmonton,  AB, 
Canada:  Northern  Forestry  Centre. 

Nei,  M.  1972.  Genetic  distance  between  populations. 
American  Naturahst.  106:  283-292. 

Nei,  M.  1975.  Molecular  population  genetics  and  evolution. 
New  York:  American  Elsevier.  288  p. 

Nei,  M.  1977.  F-statistics  and  analysis  of  gene  diversity  in 
subdivided  populations.  Annals  of  Human  Genetics.  41: 
225-233. 

Nei,  M.  1978.  Estimation  of  average  heterozygosity  and 
genetic  distance  from  a  small  niunber  of  individuals. 
Genetics.  89:  583-590. 


29 


Nei,  M.  1987a.  Genetic  distance  and  molecular  phylogeny.  In: 
Ryman,  N.;  Utter,  F.,  eds.  Population  genetics  and  fishery 
management.  Seattle,  WA:  University  of  Washington 
Press:  193-223. 

Nei,  M.  1987b.  Molecular  evolutionary  genetics.  New  York: 
Columbia  University  Press.  512  p. 

Nei,  M.;  Miller,  J.  C.  1990.  A  simple  method  for  estimating 
average  nvunber  of  nucleotide  substitutions  within  and 
between  populations  from  restriction  data.  Genetics.  125: 
873-879. 

Politov,  D.  V.  1989.  Allozyme  polymorphism,  genetic  differ- 
entiation, and  mating  system  in  Siberian  stone  pine,  Pinus 
sibirica  Du  Tour.  Moscow:  Institute  of  General  Genetics, 
Russian  Academy  of  Sciences.  190  p.  Dissertation  Thesis. 
[In  Russian]. 

Politov,  D.  v.;  Krutovskii,  K.  V.;  Altukhov,  Yu.  P.  1989. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  III.  Linkage  relationships  among  isozyme  loci. 
Genetika.  25(9):  1606-1618.  [In  Russian].  [Enghsh  trans- 
lation published  by  Plenum  Publishing  Corp.  in  Soviet 
Genetics;  1990  March:  1053-1063]. 

Politov,  D.  v.;  Krutovskii,  K.  V.  1990.  Genetic  variability 
in  Siberian  stone  pine,  Pinus  sibirica  Du  Tour.  V.  Analy- 
sis of  mating  system.  Genetika.  26(11):  2002-2011.  [In 
Russian].  [English  translation  published  by  Plenum  Pub- 
Hshing  Corp.  in  Soviet  Genetics;  1991  May:  1309-1316]. 

Politov,  D.  v.;  Krutovskii,  K  V.;  Altukhov,  Yu.  P.  1992.  Iso- 
zyme loci  characteristics  of  gene  banks  of  populations 
of  cedar  pines.  Genetika.  28(1):  93-114.  [In  Russian]. 
[English  translation  published  by  Plenum  Publishing 
Corp.  in  Soviet  Genetics;  1992  July:  76-95]. 

Rogers,  J.  S.  1972.  Measures  of  genetic  similarity  and  ge- 
netic distance.  In:  Studies  in  Genetics  VII.  Publ.  7213. 
Austin,  TX:  University  of  Texas:  145-153. 

Rohlf,  F.  J.  1988.  NTSYS-pc:  numerical  taxonomy  and  mul- 
tivariate analysis  system.  New  York:  Exeter  Publishing 
Ltd. 

Saitou,  N.;  Nei,  M.  1987.  The  neighbor-joining  method:  a 
new  method  for  reconstructing  phylogenetic  trees.  Molecu- 
lar Biology  and  Evolution.  4:  406-425. 

Schuster,  W.  S.  F.;  Mitton,  J.  B.  1991.  Relatedness  within 
clusters  of  bird-dispersed  pine  and  the  potential  for  kin 
interactions.  Heredity.  67:  41-48. 

Shaw,  G.  R.  1914.  The  genus  Pinus.  Publ.  Arnold  Arboretimi 
No.  5.  Cambridge,  IVLA:  Riverside  Press.  96  p. 

Shurkhal,  A.;  Podogas,  A.;  Zhivotovsky,  L.  1991a.  Genetic 
differentiation  of  18  pine  species  on  allozyme  loci,  genus 
Pinus,  subgenus  Strobus  and  Pinus.  Doklady  Akademii 
Nauk  SSSR.  316(2):  484-488.  [In  Russian].  [English  trans- 
lation published  by  Plenum  Publishing  Corp.  in  Doklady 
Biological  Sciences;  1991]. 

Shurkhal,  A.;  Podogas,  A.;  Zhivotovsky,  L.  1991b.  Phyloge- 
netic analysis  of  genus  Pinus  by  allozyme  loci:  genetical 
differentiation  of  subgenus.  Genetika.  27(5):  1193-1205. 
[In  Russian].  [English  translation  published  by  Plenum 
Publishing  Corp.  in  Soviet  Genetics;  1991  November]. 

Shurkhal,  A.;  Podogas,  A.;  Zhivotovsky,  L.  1992.  Allozyme 
differentiation  in  the  genus  Pinus.  Silvae  Genetica.  41(2): 
105-109. 

Sneath,  P.  H.  A.;  Sokal,  R.  R.  1973.  Numerical  taxonomy. 
San  Francisco:  W.  H.  Freeman:  230-234. 


Steinhoff,  R.  J.;  Joyce,  D.  G.;  Fins,  L.  1983.  Isozyme  varia- 
tion in  Pinus  monticola.  Canadian  Journal  of  Forest  Re- 
search. 13(6):  1122-1131. 

Strauss,  S.  H.;  Doerksen,  A.  H.  1990.  Restriction  fragment 
analysis  of  pine  phylogeny.  Evolution.  44(4):  1081-1096. 

Strauss,  S.  H.;  Bosquet,  J.;  Hipkins,  V.  D.;  Hong,  Y.-P. 
1992.  Biochemical  and  molecular  genetic  markers  in  bio- 
systematic  studies  of  forest  trees.  New  Forests.  6: 125-158. 

Swofford,  D.  L.;  Selander,  R.  B.  1981.  BIOSYS-1:  a  FOR- 
TRAN program  for  the  comprehensive  analysis  of  elec- 
trophoretic  data  in  population  genetics  and  systematics. 
Journal  of  Heredity.  72:  281-283. 

Szmidt,  A.  E.  1991.  Phylogenetic  and  applied  studies  on 
chloroplast  genome  in  forest  conifers.  In:  Hattemer,  H.  H.; 
Fineschi,  S.;  Cannata,  F.;  Malvolti,  E.  V.,  eds.  Biochemi- 
cal markers  in  the  population  genetics  of  forest  trees. 
The  Hague,  The  Netherlands:  SPB  Academic  Publishing: 
185-196. 

Szmidt,  A.  E.;  Sigurgeirsson,  A.;  Wang,  X.  R.;  Hdllgren, 
J.  E.;  Lindgren,  D.  1988.  Genetic  relationships  among 
Pinus  species  based  on  chloroplast  DNA  pol3miorphism. 
In:  H^lgren,  J.-E.,  ed.  Frans  Kempe  Sympositun  in  Umeli 
on  molecular  genetics  of  forest  trees:  proceedings;  1988; 
Umed,  Sweden.  Ume&,  Sweden:  Swedish  University  of 
Agricultural  Sciences:  33-47. 

Tomback,  D.  F.  1983.  Nutcrackers  and  pines:  coevolution  or 
coadaptation?  In:  Nitecki,  N.  H.,  ed.  Coevolution.  Chicago: 
University  of  Chicago  Press:  179-223. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Coevo- 
lution of  whitebark  pine  and  nutcrackers:  implications  for 
forest  regeneration.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposiimi  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
movmtain  Research  Station:  118-129. 

Tomback,  D.  F.;  Holtmeier,  F.  K.;  Mattes,  H.;  Carsey,  K.  S.; 
Powell,  M.  L.  1992.  Tree  clusters  and  growth  form  distri- 
bution in  Pinus  cembra,  a  bird-dispersed  pine.  Denver, 
CO:  University  of  Colorado  at  Denver,  Department  of 
Biology.  [Working  draft]. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Wheeler,  N.  C;  Guries,  R.  P.  1982.  Population  structure, 
genetic  diversity  and  morphological  variation  in  Pinus 
contorta  Dougl.  Canadian  Journal  of  Forest  Research. 
12(3):  595-606. 

Wheeler,  N.  C;  Guries,  R.  P.  1987.  A  quantitative  measure 
of  introgression  between  lodgepole  and  jack  pines.  Cana- 
dian Journal  of  Botany.  65:  1876-1885. 

Wheeler,  N.  C;  Guries,  R.  P.;  O'Malley,  D.  M.  1983.  Bio- 
systematics  of  the  genus  Pinus,  subsection  Contortae. 
Biochemical  Systematics  and  Ecology.  11:  333-340. 

Yeh,  F.  C.  H.;  Cheliak,  W.  M.;  Dancik,  B.  P.;  Illingworth,  K; 
Trust,  D.  C;  Pryhitka,  B.  A.  1985.  Population  differentia- 
tion in  lodgepole  pine,  Pinus  contorta  ssp.  latifolia:  a  dis- 
criminant analysis  of  allozyme  variation.  Canadian  Jour- 
nal of  Genetics  and  Cytology.  27(2):  210-218. 


30 


COEVOLUTIONAL  ASPECTS  OF  STONE 
PINES  AND  NUTCRACKERS 

Hermann  Mattes 


Abstract — ^Nutcrackers  (Nucifraga  caryocatactes  and  N. 
Columbiana)  have  developed  a  strong  mutualistic  relationship  to 
stone  pines.  Seeds  of  stone  pines  are  the  most  important  food  for 
at  least  9  months  of  the  year  and  for  raising  the  young.  In  addi- 
tion to  special  adaptations  on  gathering,  transporting,  caching, 
and  finding  again  the  hoarded  seeds,  the  whole  annual  cycle  of 
the  nutcracker's  life  (time  of  breeding  and  moulting),  its  mating 
system,  and  its  habitat  use  are  adjusted  to  the  use  of  pine  seeds. 


Within  the  family  of  jays,  magpies,  and  crows  (Cor- 
vidae)  many  characteristics  of  preadaptive  value  for  seed 
dispersal  occur  (Tomback  and  Linhart  1990;  Turcek  and 
Kelso  1968).  Three  corvid  species  have  evolved  a  strong 
mutualistic  system  with  pines:  Clark's  nutcracker  (Nuci- 
fraga Columbiana)  and  the  pinyon  jay  (Gymnorhinus 
cyanocephalus)  in  North  America  and  the  Exiropean  nut- 
cracker {Nucifraga  caryocatactes)  in  Eurasia.  Compre- 
hensive studies  on  food  regime  and  seed  dispersal  were 
carried  out  by  many  authors  (Bibikov  1948;  Crocq  1990; 
Hayashida  1982,  these  proceedings;  Marzlviif  and  Balda 
1992;  Mattes  1978;  Reijmers  1959;  Saito  1983;  Swanberg 
1951;  Tomback  1977;  Vander  Wall  and  Balda  1977,  1981). 
Coevolution  in  corvids  and  pines  does  not  only  refer  to 
feeding  and  caching  behavior,  but  the  whole  life  cycle  is 
adjusted  to  the  use  of  the  special  food  resource.  This  pa- 
per describes  the  process  for  the  European  nutcracker. 

THE  EUROPEAN  NUTCRACKER 

The  annual  life  cycle  (fig.  1)  of  the  nutcracker  is  best 
looked  at  fi'om  August  onward.  Unripe  seeds  of  the  Swiss 
stone  pine  (cembra  pine,  Pinus  cemhra)  are  eaten  by  the 
nutcracker  throughout  August  or  even  in  July;  thus,  the 
bird  will  not  miss  the  point  when  the  seeds  are  fully  de- 
veloped. At  this  time  the  resin  flow  in  the  cone  stops,  and 
cone  scales  get  loose.  Now  the  seeds  can  be  easily  taken 
out  of  the  cone,  and  the  nutcracker  begins  gathering  and 
caching  seeds  for  winter  supply.  Pine  seed  gathering  will 
last  until  all  seeds  have  been  harvested  and  almost  no 
cones  are  left  on  the  trees.  Only  twice  within  19  years  of 
investigation  in  the  Engadine  a  cone  crop  lasted  until  the 
beginning  of  the  following  spring. 

In  the  Alps,  nutcrackers  carry  seeds  of  cembra  pine  over 
distances  up  to  15  km  (including  differences  in  altitude  of 
at  least  700  m)  back  to  their  territories.  They  carry  45 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Hermann  Mattes  is  Professor  of  Landscape  Ecology  and  Geography, 
Department  of  Geography,  University  of  Miinster,  Germany. 


pine  seeds  per  flight  on  an  average;  however,  a  maximum 
load  of  134  cembra  pine  seeds  has  been  found  in  the  sub- 
lingual pouch  of  a  nutcracker.  Main  morphological  and 
ethological  adaptations  for  seed  use  are: 

•  a  long  and  pointed  bill  (fig.  2)  that  enables  the  bird 

to  take  seeds  out  of  the  cones  as  one  would  with  twee- 
zers (most  other  corvids  have  a  hooked  bill); 

•  a  rist  in  the  lower  mandible  allows  the  bird  to  fix  and 
crack  pine  seeds,  supported  by  a  short-cut  tongue; 

•  the  subhngual  pouch  allows  carrying  a  large  quantity 
of  seeds  (fig.  3)  without  disturbing  other  functions  of 
the  bill; 

•  ability  to  test  seed  quality  by  means  of  seed  color  and 
resonance  (by  so-called  bill-clicking). 

SEED  CACHES 

The  cached  seeds  provide  the  main  food  source  for  the 
nutcracker  from  November  to  April,  and  to  a  lesser  degree 
until  July.  A  single  nutcracker  stores  more  than  25,000 
seeds.  As  yet  it  has  not  been  clearly  shown  how  the  nut- 
cracker can  retrieve  its  seed  caches.  However,  as  can  be 
concluded  fi'om  observations  in  the  field  and  in  the  labora- 
tory, nutcrackers  apparently  remember  each  single  cache 
(Tomback  1982,  and  others).  Further,  there  are  no  spe- 
cific features  of  caches,  although  nutcrackers  obviously 
prefer  to  deposit  seeds  near  border  lines  and  any  changes 
of  surface  structure.  Experiments  with  the  Clark's  nut- 
cracker by  Kamil  and  Balda  (1988)  and  Vander  Wall 
(1982)  showed  that  nutcrackers  orient  (navigate)  opti- 
cally, as  expected,  and  additionally  with  a  magnetic  sense. 
Understanding  cache  findings,  however,  requires  more  de- 
tailed information,  as  nutcrackers  detect  caches  even  un- 
der a  thick  cover  of  snow.  Balda  and  Kamil  (1989)  sug- 
gest an  overall  fuzzy  optical  navigation  for  finding  caches. 
This  wovild  allow  the  birds  to  locate  a  cache  even  if  there 
were  distinctive  changes,  for  example,  in  height  of  snow 
cover,  broken  branches,  or  fallen  trees.  It  is  still  an  open 
question  how  the  bird  can  find  caches  under  a  snow  cover 
of  20  cm  and  higher,  getting  very  directly  to  the  hoarded 
seeds.  Snow  tunnels  of  130  cm  (Burckhardt  1958)  and 
about  300  cm  (R.  Stem,  personal  communication)  were 
reported,  and — what  is  most  surprising — the  nutcrackers 
did  find  the  cached  seeds.  Crocq  (1990)  found  the  nut- 
crackers leaving  their  subalpine  habitat  when  snow  cover 
reached  170  cm  on  imdistixrbed  surfaces. 

It  shoiild  be  emphasized  that  nutcrackers  dig  success- 
fully for  their  caches  throughout  winter  imtil  early  spring. 
That  guarantees  sufficient  food  for  the  brood  during  that 
difficult  time  of  the  year.  Nutcrackers  must  remember 
not  only  the  sites  of  the  caches  but  additionally  those  that 
already  have  been  used.  This  doubles  the  efforts  of 


31 


> 

twmm 


11 


Jan.     Feb.    Mar.      Apr.    May    June     July     Aug.    Sept.     Oct.    Nov.  Dec. 

Month 

Figure  1 — Annual  activities  of  the  nutcracker  in  the  Alps  for  years  with  a  medium  cone  crop.  1  -3,  feeding: 
1 ,  seed  harvest  and  transport;  2,  use  of  cached  seeds;  3,  consumption  of  animal  food,  for  example,  spi- 
ders, insects;  4,  winter  period;  5-7  breeding  period:  5,  courtship  and  nest  building;  6,  time  with  clutch  and 
nestlings;  7,  family  groups  with  fledged  young;  8-9,  moulting:  8,  moult  of  adults;  9,  moult  of  fledglings; 
10-11,  movements:  10,  dismigration;  11,  migration. 


memory.  There  is  some  evidence  that  nutcrackers  use 
caches  from  the  penultimate  seed  crop,  which  may  be  im- 
portant in  case  a  cone  crop  fails.  In  years  of  small  seed 
crops,  nutcrackers  often  recache  the  seeds.  Temporary 
caches  are  made  in  cases  of  strong  competition  for  seeds 
(when  great  nimabers  of  nutcrackers  and  squirrels  concen- 
trate on  a  few  cone-bearing  trees).  In  temporary  caches, 
nutcrackers  mostly  hide  cones. 

BREEDING  AND  HABITAT 

Nutcrackers  start  breeding  in  March  (fig.  4).  At  this 
time  mean  daily  temperature  is  under  0  °C.  Early  breed- 
ing is  necessary  because  the  next  generation  has  to  be 
ready  for  harvesting  seeds  in  August  to  ensure  the  winter 
food  supply.  Breeding  takes  about  18  days,  the  nestlings 
fledge  when  24  to  28  days  old,  and  there  is  an  extraordi- 
narily long  time  of  parental  care  for  the  fledglings — until 
they  are  about  100  days  old  (Oberauer  1991;  Rudat  1978). 
Extensive  care  by  the  parents  is  needed  because  of  scarcity 


Figure  2 — Inside  the  nutcracker's  bill:  short 
split  tongue,  underneath  opening  of  sublin- 
gual pouch  with  peanut  in  the  entrance  part, 
rist  in  the  lower  mandible. 


32 


Figure  3 — Nutcracker  with  filled  sublingual 
pouch.  Morteratsch,  Engadin;  September 
1975. 


of  food  up  to  mid-June,  and  to  provide  training  for  caching 
and  digging  for  seeds. 

Courtship  behavior  starts  in  the  end  of  February  or 
early  March.  Synchronizing  with  the  partner  is  of  ex- 
treme importance  for  successful  breeding.  By  February 
the  nutcrackers'  so-called  ceremonial  gatherings  begin 
and  then  can  be  frequently  observed  (Swanberg  1956). 
Up  to  12  or  more  nutcrackers  from  neighboring  and  some- 
times far  territories  meet  in  conspicuous  noisy  groups.  An 
individually  marked  female  nutcracker  that  had  spent  the 
winter  outside  its  territory  came  back  late — March  17.  I 
discovered  her  fighting  for  the  territory  while  a  ceremo- 
nial gathering  was  going  on.  She  succeeded  in  reoccupy- 
ing  her  former  place,  and  raised  two  nestlings  after  hav- 
ing a  three-egg  clutch  with  one  unfertile  and  half-sized 
egg.  Neighbors  know  each  other  individually.  So,  except 
in  artifically  created  feeding  places,  fights  are  very  rare. 

Nutcracker  territories  are  organized  in  a  different  way 
than  those  of  other  passerines.  Only  the  area  very  close 
to  the  nest  is  defended  against  intruders  and  thus  can  be 
called  "territory"  in  the  usual  sense.  In  a  broader  sense, 
the  territory  comprises  the  area  of  food  caches  and  is 
about  5  to  12  ha  large  according  to  habitat.  These  "terri- 
tories" can  overlap  and  are  not  defended  because  cached 
seeds  are  only  accessible  to  the  very  same  nutcracker.  In 
addition,  all  places  regularly  visited  for  seed  gathering 
and  transport  should  be  regarded  as  home  range.  Habi- 
tats of  the  nutcracker  are  characterized  by  evergreen  co- 
nifers. In  such  forests  sheltered  nesting  places  are  avail- 
able, and  because  snow  cover  is  low  under  dense  tree 
canopies  cached  seeds  can  be  easily  recovered. 


Nutcrackers  are  adapted  in  many  ways  for  breeding 
under  winter  conditions.  The  nest  is  heavily  built  with  a 
layer  of  dry  decayed  wood  for  insulation  against  low  tem- 
perature. Beginning  with  the  first  egg,  the  parents  must 
warm  the  clutch  without  long  interruptions.  Breeding 
starts  after  the  last  egg  has  been  laid,  and  egg  tempera- 
ture is  raised  from  about  25  °C  to  breeding  conditions  of 
about  38  °C.  Different  from  most  corvids,  both  adults 
breed.  The  male  develops  a  large  breeding  patch,  too. 
The  female's  breeding  time  is  somewhat  higher  during  the 
day,  and  it  also  stays  on  the  nest  at  night.  The  nestlings 
are  fed  with  pine  seeds  from  their  first  day  of  life.  Addi- 
tional food  consists  of  spiders,  and,  if  available,  insects. 
Nutrient  analysis  in  seeds  of  whitebark  pine  {Pinus 
albicaulis)  by  Lanner  and  Gilbert  (these  proceedings) 
showed  a  high  nutritive  quality  in  caloric  value,  mineral 
nutrients,  fatty  acids,  and  amino  acids.  Should  this  apply 
also  to  cembra  pine  seeds,  it  is  easy  to  understand  why 
first-day  nestlings  can  live  mainly  on  a  seed  diet. 

For  successful  overwintering  and  reproduction,  suffi- 
cient numbers  of  seed  caches  must  be  available  to  both 
parents.  First-year  birds  dismigrate  in  August  and  estab- 
lish their  own  territory  when  seed  caching  starts.  First- 
year  birds  that  are  not  fully  developed  until  August  will 
have  a  minor  chance  to  survive  winter.  A  few  first-year 
birds  leave  their  native  area  in  September  and  migrate  in 
southern  directions  (Mattes  and  Jenni  1984).  Adult  nut- 
crackers are  highly  sedentary.  Nonbreeding  individuals 
might  have  action  ranges  within  or  between  the  territo- 
ries of  breeding  birds.  This  popiilation  reserve  can  com- 
pensate for  winter  losses  of  adult  birds. 

In  the  "Stazer  Wald"  I  observed  a  third  bird  helping  a 
pair  build  their  nest;  unfortunately,  no  further  observa- 
tion was  possible.  Similar  to  several  jay  species,  helping 
in  breeding  may  be  an  evolutional  feature  in  the  nut- 
cracker (Woolfenden  and  Fitzpatrick  1984).  Nutcrackers 
live  monogamously  for  life,  as  can  be  concluded  from  indi- 
vidually marked  birds  (Mattes  1978;  Swanberg  1956).  In 


Figure  4 — A  nutcracker  breeding,  the  nest 
completely  covered  by  snow.  March  1 9, 
1975,  Engadin. 


33 


the  Engadine,  the  mortality  rate  in  adult  birds  was  84  per- 
cent (Mattes,  in  preparation)  on  an  average.  The  oldest 
birds  found  were  at  least  15.  Keeping  territories  and  part- 
ners for  life  supports  synchronization  of  the  parents,  and 
individual  knowledge  of  neighbors  minimizes  social  com- 
petition. Social  hierarchy  is  evident  and  most  conspicu- 
ous at  artifical  feeding  places.  Females  take  over  the  so- 
cial rank  of  their  mates  (Grabher-Meyer  1991). 

In  passerines,  moult  normally  takes  place  after  the 
breeding  season.  Prolonged  care  of  the  fledglings  and 
beginning  of  seed  hoarding  in  August  would  not  leave 
time  for  movdting  in  late  summer.  Actually,  moult  starts 
in  March  with  change  of  remiges  and  tail  feathers,  and 
ends  in  July  when  body  feathers  are  changed.  Moult  com- 
pletely parallels  breeding  season;  the  double  physiological 
effort  causes  a  very  long  moulting  period.  Body  feathers 
are  moulted  relatively  quickly  during  the  warmest  part  of 
the  year. 

REACTION  TO  CHANGES 

Predation  on  pine  seeds  has  influenced  the  whole  life 
cycle  of  the  nutcracker.  Many  features  common  to  the 
corvids  are  particularly  developed  in  the  nutcracker.  Still 
there  is  a  broad  and  flexible  reaction  to  environmental 
changes.  This  is  most  apparent  during  the  invasions  of 
Siberian  nutcrackers  (N.  c.  macrorhynchos)  into  western 
Europe.  Diet  and  feeding  as  well  as  habitat  use  vary 
widely.  Nevertheless,  most  of  the  invading  birds  very 
likely  die.  Some,  however,  return  to  their  breeding  ground 
either  in  the  current  autumn  or  in  the  next  spring.  Nimaer- 
ous  birds  ringed  in  the  Aland  archipelago  (Finland)  in 
August  1968  were  recaptured  a  few  weeks  later  in  the 
East;  three  birds  were  found  beyond  the  Urals  at  a  maxi- 
mum distance  of  almost  3,000  km  (Zink  1981).  Invaders 
stayed  at  wintering  places  in  Bielefeld,  Germany,  when 
Siberian  pine  seeds  (Pinus  sibirica)  were  offered  ad 
libidimi  (Conrads  and  Balda  1979). 

Outside  the  natural  range  of  cembra  pine  (Central  Alps, 
Tatra,  and  a  few  locations  in  the  Carpathian  mountains) 
the  European  nutcracker  normally  feeds  on  hazel  nuts 
iCorylus  avellana).  Slight,  but  statistically  significant, 
differences  in  bill  size  are  apparent  in  central-alpine 
populations.  Nutcrackers  from  the  Engadine  and  migrat- 
ing birds  at  Col  de  Bretolet  in  southwestern  Switzerland 
show  a  narrower  lower  mandible  and  a  longer  bill  com- 
pared with  populations  from  southern  Germany  and  from 
northern  Europe  (Mattes  1978;  Mattes  and  Jenni  1984). 
Also,  there  are  distinct  differences  in  bill  size  in  popula- 
tions of  A^.  c.  macella  from  Szetchuan  and  from  Sikkim 
(Mattes,  unpublished).  Generally,  there  seems  to  be  a 
correlation  of  width  of  lower  mandible  and  size  of  main 
food.  The  distance  between  the  branches  of  the  lower 
mandible  limits  the  size  of  the  entrance  to  the  sublingual 
pouch.  As  the  main  food  of  most  Asian  populations  of  the 
nutcracker  is  not  yet  sufficiently  known,  bill  size  and  bill 
structure  cannot  yet  be  fully  explained.  Furthermore,  bill 
size  in  some  subspecies  may  reflect  feeding  conditions  at 
former  times. 


REFERENCES 

Balda,  Russel  P.;  Kamil,  Alan  C.  1989.  A  comparative 

study  of  cache  recovery  by  three  corvid  species.  Animal 

Behavior.  38:  486-495. 
Bibikov,  D.  I.  1948.  On  the  ecology  of  the  nutcracker. 

Trudy  Pechoro-Ilychkoge  Gosud.  Zapovednik.  4:  89-112. 
Biu-ckhardt,  D.  1958.  Vom  Vorratanlegen  des  Tannen- 

hahers.  Biindnerwald.  11:102-114. 
Conrads,  Klaus;  Balda,  Russell  P.  1979.  Uberwinter- 

imgschancen  Sibirischer  Tannenhaher  {Nucifraga 

caryocatactes  macrorhynchos)  im  Invasionsgebiet. 

Bericht  des  Naturwissenschaftlichen  Vereins  Bielefeld. 

24:  115-137. 

Crocq,  Claude.  1990.  Le  casse-noix  mouchete  {Nucifraga 
caryocatactes).  Paris:  Lechevalier/R.  Chabaud.  326  p. 

Grabher-Meyer,  Arno.  1991.  Beobachtungen  zur  Brut- 
biologie  einer  kleinen  Population  des  Tannenhahers 
{Nucifraga  caryocatactes)  bei  Praxmar,  Sellrain. 
Diplomarbeit,  Universitat  Innsbruck,  Austria. 

Kamil,  Alan  C;  Balda,  Russell  P.  1988.  Spatial  memory 
in  Clark's  nutcracker  {Nucifraga  columbiana).  Acta  XIX 
Congressus  Internat.  Ornithol.,  Ottawa  1986.  Vol.  II: 
2100-2111. 

Lanner,  Ronald  M.  1982.  Adaptations  of  whitebark  pine 
for  seed  dispersal  by  Clark's  nutcracker.  Canadian 
Journal  of  Forest  Research.  12(2):  391-402. 

Marzluff,  John  M.;  Balda,  Russ  P.  1992.  The  pinyon  jay: 
behavioral  ecology  of  a  colonial  and  cooperative  corvid. 
London:  Poyser.  317  p. 

Mattes,  Hermann  R.  1978.  Der  Tannenhaher  im  Engadin. 
Studien  zu  seiner  Okologie  und  Funktion  im  Arven- 
wald.  Miinstersche  Geographische  Arbeiten  2. 
Paderbom:  Schoning.  87  p. 

Mattes,  Hermann  R.;  Jenni,  Lukas.  1984.  Ortstreue  und 
Zugbewegungen  des  Tannenhahers  {Nucifraga  caryo- 
catactes) im  Alpenraima  und  am  Randecker  Maar/ 
Schwabische  Alb.  Ornithologischer  Beobachter.  81: 
303-315. 

Oberauer,  Eva.  1991.  Zur  Postembryonalentwicklimg  des 
Tannenhahers  {Nucifraga  caryocatactes  L.).  Beobach- 
tungen im  Alpenzoo  Innsbruck.  Diplomarbeit,  Uni- 
versitat Innsbruck,  Austria.  140  p. 

Reijmers,  N.  F.  1959.  Birds  of  the  cedar-pine  forests  of 
south-central  Siberia  and  their  role  in  the  life  of  the 
cedar-pine.  Trudy  biolog.  Inst,  sibirsk.  Otdel.  Akad. 
Nauk  SSSR.  5:  121-166. 

Rudat,  Volker;  Rudat,  Wolf.  1978.  Zum  Verhalten  von 
Tannenhaherfamilien  {Nucifraga  caryocatactes  L.)  in 
der  Zeit  vom  Ausfliegen  der  Jungen  bis  zu  deren 
Selbstandigwerden.  Zoologisches  Jahrbuch  fur 
Systematik.  105:  386-398. 

Saito,  Shin-ichiro.  1983.  On  the  relations  of  the  caching 
by  animals  on  the  seed  germination  of  Japanese  stone 
pine,  Pinus  pumila  Regel.  Bulletin  of  the  Shiretoko 
Museum.  5:  23-40. 

Swanberg,  Peer  O.  1951.  Food  storage,  territory  and  song 
in  the  thickbilled  nutcracker.  Acta  X  Congressus 
Internat.  Ornitholog.  Uppsala  1950:  545-554. 

Swanberg,  Peer  O.  1956.  Territory  in  the  thickbilled  nut- 
cracker {Nucifraga  caryocatactes).  Ibis.  98:  412-419. 


34 


Tomback,  Diana  F.  1977.  Foraging  strategies  of  Clark's 

nutcracker.  The  Living  Bird.  17:  123-161. 
Tomback,  Diana  F.  1982.  How  nutcrackers  find  their  seed 

stores.  The  Condor.  82:  10-19. 
Tomback,  Diana  F.;  Linhart,  Yan  B.  1988.  The  evolution 

of  bird-dispersed  pines.  Evolutionary  Ecology.  4: 

185-219. 

Vander  Wall,  Stephen  B.  1982.  An  experimental  analysis 
of  cache  recovery  in  Clark's  nutcracker.  Animal 
Behavior.  30:  84-94. 

Vander  Wall,  Stephen  B.;  Balda,  Russ  P.  1977.  Coadapta- 
tions  of  the  Clark's  nutcracker  and  the  pinyon  pine  for 


efficient  seed  har\'est  and  dispersal.  Ecological  Mono- 
graphs. 47(1):  27-37. 
Vander  Wall,  Stephen  B.;  Balda,  Russ  P.  1981.  Ecology 
and  evolution  of  food  storage  behavior  in  conifer- 
seed-caching  Corvids.  Zeitschrift  fiir  Tierpsychologie. 
56:  217-242. 

Woolfenden,  D.  E.;  Fitzpatrick,  W.-J.  1984,  The  Florida 
scrub  jay.  Princeton,  NJ:  Princeton  University  Press. 

Zink,  Gerhard.  1981.  Der  Zug  europaischer  Zugvogel.  3. 
Liefenmg.  Moggingen:  Vogelzug-Verlag. 


35 


ALLOZYME  POLYMORPHISM, 
HETEROZYGOSITY,  AND  MATING 
SYSTEM  OF  STONE  PINES 

Dmitri  V.  Politov 
Konstantin  V.  Krutovskii 


Abstract— Population  genetic  structure  and  the  mating  system  of 
stone  pines  (genus  Pinus,  subsection  Cembrae)  have  been  studied 
using  isozymes.  Observed  were  a  tendency  to  heterozygote  defi- 
ciency among  embryos  (obviously,  a  consequence  of  self-pollination 
with  the  estimated  fi-equency  of  0.11-0.30)  and  a  slight  heterozy- 
gote excess  in  adult  trees  (probably  caused  by  selection  against 
inbred  progeny  and  balanced  selection). 


Electrophoretic  analysis  of  allozjone  polymorphism  is 
one  of  the  most  useful  methods  for  study  of  genetic  and  evo- 
lutionary processes  in  natural  populations  and  taxonomic 
and  phylogenetic  relationships  of  forest  tree  species  (Adams 
and  others  1992;  Fineschi  and  others  1991;  MuUer-Starck 
and  Ziehe  1991).  Conifers  are  particularly  attractive,  from 
a  genetic  point  of  view,  because  of  their  haploid  megaga- 
metophyte  (endosperm).  Electrophoresis  of  both  haploid 
and  diploid  (embryo)  seed  tissues  permits  us  to  distinguish 
maternal  and  paternal  contributions  into  the  progeny  geno- 
type. This  makes  it  possible  to  estimate  allele  frequencies 
separately  in  both  male  and  female  gametes  pools  and  to 
study  segregating  ratios  directly  by  analyzing  endosperm 
haplotypes,  which  correspond  to  haplotypes  of  female  ga- 
metes. The  possibility  of  observing  meiotic  products  di- 
rectly through  study  of  haploid  endosperms  simplifies  analy- 
sis of  allozyme  inheritance,  genetic  Unkage  of  isozyme  loci, 
and  the  mating  system  of  conifers. 

In  contrast  to  many  widespread  European  and  North 
American  conifer  species,  stone  pines  have  been  studied 
using  isozymes  only  recently  (Groncharenko  and  others  1988; 
Krutovskii  and  Politov  1992;  Krutovskii  and  others  1987, 
1988, 1989, 1990;  Podogas  and  others  1991;  Politov  1989; 
PoUtov  and  Krutovskii  1990;  Politov  and  others  1989,  1992; 
Szmidt  1982).  Meanwhile,  the  need  for  such  data  is  appar- 
ent both  for  practical  forestry  applications  and  for  purposes 
of  genetic  diversity  conservation.  As  a  major  forest-forming 
tree  species,  stone  pines  determine,  to  a  large  extent,  the 
state  of  ecosystems  over  extensive  areas.  Their  genetic  vul- 
nerability thus  threatens  the  existence  of  a  unique  forest 
type. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11, 1992. 

Dmitri  V.  Politov  is  Research  Scientist  and  Konstantin  V.  Krutovskii 
is  Senior  Research  Scientist,  Laboratory  of  Population  Genetics,  Russian 
Academy  of  Sciences,  Gubkin  Str.  3, 117809  GSP-1  Moscow,  Russia. 


Four  of  the  five  stone  pine  species  traditionally  united 
in  subsection  Cembrae  of  genus  Pinus  (Critchfield  and 
Little  1966)  are  found  in  Russia  and  neighboring  countries: 
Siberian  stone  pine  (Pinus  sibirica  Du  Tour),  European  or 
Swiss  stone  pine  (P.  cembra  L.),  Korean  pine  {P.  koraiensis 
Siebold  et  Zucc),  and  mountain  or  Japanese  stone  pine 
(P.  pumila  [Pall.]  Kegel).  All  of  them  are  very  important 
economically. 

This  paper  reports  the  main  results  of  long-term  popula- 
tion genetic  studies  on  these  species  using  isozyme  analy- 
sis. Some  population  structure  parameters  also  have  been 
estimated  for  one  population  of  whitebark  pine  (P.  albicaulis 
Engelm.),  the  only  representative  of  stone  pines  in  the 
Western  Hemisphere. 

MATERIALS  AND  METHODS 

Sampling  Sites— Seeds  collected  in  natural  populations 
served  as  a  source  of  laboratory  material.  Pinus  sibirica 
seeds  were  collected  from  11  localities.  Samples  desig- 
nated as  Filin  Klyuch  (F),  Mutnaya  Rechka  (M),  Maljd 
Kebezh  (MK),  Sobach'ya  Rechka  (SR),  and  Listvyanka  (L) 
originate  from  the  Western  Sayan  (Ermakovskii  District, 
Krasnoyarsk  Territory);  Yayl'u  (YA)  from  the  Altfd  Moun- 
tains (Gomo-Altaiskaya  Autonomous  Region);  Smokotino 
(S),  Zorkal'tsevo-1  (Z-1),  and  Zorkal'tsevo-3  (Z-3)  from  Tom- 
skaya  Region;  Turukhan  (TU)  from  the  north  of  Krasnoy- 
arsk Territory  (Turukhanskii  District);  and  N'alino  (NYA) 
from  the  Khanty-Mansiiskii  Autonomous  District  (Tjoimen- 
skaya  Region).  Besides,  one  locality  of  P.  cembra,  Ust'- 
Chornaya  (UCH)  from  the  Eastern  Carpathians  (Ukraine); 
three  localities  of  P.  pumila  from  the  north  of  Kamchatka 
Peninsula  designated  as  Grustnyi  (GR),  Uttyveem  (UT), 
and  Kamenskoye  (K);  and  three  localities  of  P.  koraiensis 
from  Khabarovsk  Territory  designated  Malokhekhtsirskoye 
(MKH)  and  Khor  (KHO),  and  from  Primorskii  Territory 
(Sikhote-Alin'  (SA))  have  been  also  studied.  The  sample 
of  P.  albicaulis  originates  from  the  Livingstone  Falls  (LF) 
population  (southwestern  Alberta,  Canada). 

Electrophoresis  and  Data  Processing— No  less  than 
seven  endosperms  and  embryos  per  tree  have  been  studied 
using  starch  gel  electrophoresis.  Totally,  536  trees  of  five 
species  (from  11  to  75  trees  per  population)  and  more  than 
3,500  embryos  (7  to  20  per  tree),  as  well  as  two  bulked  sam- 
ples of  seeds  representing  a  large  number  of  trees  (TU  and 
NYA),  have  been  analyzed.  Specimen  preparation,  buffer 
systems  used,  genetic  interpretation  of  zymograms,  and  des- 
ignations of  £dlozymes,  alleles,  and  loci  have  been  described 
elsewhere  (Krutovskii  and  others  1987;  Politov  1989). 


36 


Genotypes  of  particular  trees  were  inferred  from  segre- 
gation of  alleles  among  their  endosperms.  Then,  allele  and 
genot5T)e  frequencies  in  samples  of  both  mature  trees  and 
embryos,  as  well  as  in  the  effective  pollen  pool  (paternal 
alleles  of  embryos),  were  estimated.  Embryo  tissue  elec- 
trophoresis does  not  give  high  resolution  for  all  zones  of 
activity  and,  therefore,  calculations  based  on  embryo  sam- 
ples have  included  a  smaller  nimiber  of  loci.  Besides,  some 
activity  zones  in  individual  species  could  not  be  reliably  in- 
terpreted or  were  invariant.  Because  of  this,  specific  sets 
of  loci  used  for  particular  purposes  in  these  species  may  dif- 
fer somewhat.  More  detailed  description  of  statistic  proce- 
dures can  be  fovmd  elsewhere  (Krutovskii  and  others  1989; 
Politov  1989;  PoHtov  and  Krutovskii  1990;  Politov  and  oth- 
ers 1989). 

GENETIC  CONTROL  OF  ISOZYMES 
AND  LINKAGE  OF  LOCI 

We  have  described  inheritance  of  11  isozyme  systems 
of  stone  pines  in  our  previous  publications  (Krutovskii  and 
others  1987;  Politov  1989)  and  foimd  that  they  are  controlled 
by  at  least  26  loci.  However,  only  21  most  reliable  loci  with 
unambiguous  genetic  control  have  been  included  in  this 
study  (table  1).  Segregation  ratios  of  allelic  variants  of  in- 
dividual allozjnne  loci  observed  among  endosperms  of  het- 
erozygous trees  correspond,  with  several  exceptions,  to  the 
expected  ones  (1:1). 

Analysis  of  joint  allelic  segregation  for  pairwise  loci  com- 
binations in  Pinus  sibirica  has  revealed  genetic  linkage 
of  loci  Adh-1  and  Lap-3  (with  the  mean  recombination  fi^e- 
quency  0.275),  Adh-1  and  Pgi-2  (0.301),  and  Fe-2  and  Lap-2 
(0.309).  These  loci  belong  to  two  linkage  groups:  Lap-3 — 
Adh-1— Pgi-2  and  Lap-2— Fe-2  (Politov  and  others  1989). 
Comparison  of  these  data  with  those  available  in  literature 
shows  that  stone  pines,  as  well  as  other  representatives  of 
the  family  Pi/iacme,  are  characterized  by  extreme  conser- 
vatism of  their  genome,  since  the  same  loci  combinations 
with  nearly  the  same  values  of  recombination  fi-equency 
have  been  described  not  only  in  the  closely  related  P.  albi- 
caulis  (Furnier  and  others  1986),  but  also  in  many  hard 
pines  (see,  for  example,  Szmidt  and  Muona  1989).  Data 
presented  here  are  apparently  the  first  attempt  of  genetic 
mapping  in  one  of  the  stone  pines.  However,  one  can  as- 
sume that  the  linkage  groups  revealed  in  P.  sibirica  will 
also  be  fotmd  in  other  related  species.  Information  of  this 
kind  could  be  useful  for  breeding  programs,  for  analysis 
of  association  of  genetic  markers  with  economically  impor- 
tant characters,  and  for  mating  system  parameters  estima- 
tion by  multilocus  methods. 

POPULATION  GENETIC  STRUCTURE 

Levels  of  Intrapopulation  Variability — ^According 
to  the  gene  diversity  parameters  estimated  in  stone  pine 
populations  using  isozjone  loci,  the  level  of  genetic  vari- 
ability appears  to  be  rather  high  (table  2).  There  are  al- 
most no  differences  between  observed  heterozygosities  and 
those  expected  fi'om  corresponding  Hardy- Weinberg  pro- 
portions. The  values  of  the  latter  for  individual  species 
are:  for  P.  sibirica— from  0.140  to  0.176,  P.  cembra— 0.109, 


Table  1 — Enzymes  and  loci  analyzed 

Total 
number 


Enzyme  (abbreviation) 

of  loci 

Scored  loci 

Alcohol  dehydrogenase  (ADH) 

2 

Adh-1 ,  Adh-2 

DIaphorase  (DIA) 

3 

Dia-1 ,  Dia-2 

Fluorescent  esterase  (FE) 

3 

Fe-2 

Glutamate  dehydrogenase  (GDH) 

1 

Gdh 

G 1  utam  atoxaloacetate 

3 

Got-1 ,  Got-2,  Got-3 

transaminase  (GOT) 

Isocitrate  dehydrogenase  (IDH) 

1 

Idh 

Leucine  aminopeptidase  (LAP) 

3 

LaD-2  LaD-3 

Malate  dehydrogenase  (MDH) 

4 

Mdh-1,  Mdh-2,  Mdh-3 

Mdh-4 

Phosphoglucose  isomerase  (PGI) 

2 

Pgi-2 

Phosphoglucomutase  (PGM) 

2 

Pgm-1,  Pgm-2 

Shikimate  dehydrogenase  (SKDH) 

2 

Skdh-1 ,  Skdh-2 

P.  pumila — from  0.239  to  0.258,  P.  koraiensis — from  0.113 
to  0.148,  P.  albicaulis— 0.204. 

Such  a  high  amoimt  of  intrapopulation  variability  is  char- 
acteristic of  conifers.  We  have  summarized  data  on  allozyme 
diversity  of  30  pine  species  using  available  references 
(Politov  and  others  1992).  The  mean  number  of  alleles  per 
locus  (A)  and  the  expected  heterozygosity  (Hg),  averaged  for 
all  pine  species,  equalled  2.08  and  0.169,  correspondingly. 
Comparable  values  have  been  found  for  spruces  (A  =  2.2, 
Hg  =  0.183),  larches  (A  =  1.6,  H^  =  0.091),  and  representa- 
tives of  other  less-studied  genera. 

Factors  supposedly  promoting  a  high  level  of  genetic  diver- 
sity in  stone  pine  populations  include  such  ones,  common 
for  forest  trees,  as  predominantly  outcrossing  mating  sys- 
tem, wide  distribution,  large  effective  size  of  population, 
and  high  longevity  (Hamrick  and  Godt  1989;  Ledig  1986). 
More  specific  evolutionary  factors,  such  as  drastic  reduc- 
tion of  population  size  ("bottleneck  effect"),  may  also  affect 
the  variability  level.  Although  the  ranges  of  stone  pines, 
both  in  Eurasia  and  in  North  America,  were  reduced  diu-- 
ing  the  Quaternary  glacials  (and  also  during  the  global 
warming),  preservation  of  extensive  refugiimis  permitted 
avoidance  of  substantial  genetic  depauperization.  This 
concerns  Pinus  cembra  to  a  much  less  degree  than  other 
stone  pines. 

According  to  some  evidence,  Swiss  stone  pine  imderwent 
a  sharp  reduction  in  its  range  and  population  size  during  the 
Holocene  caused  by  the  global  warming  (Bobrov  1978  and 
references  therein).  A  small  effective  number  is  typical  of 
most  populations  of  this  species  at  present  as  well.  Swiss 
stone  pine  exists  nowadays  in  the  Alps  and  the  Carpathians 
as  small  isolated  stands  in  which  processes  of  genetic  drift 
do  not  allow  aUozyine  variability  to  be  maintained  on  a  high 
level.  This  is  a  most  probable  cause  for  a  lower  level  of  het- 
erozygosity (0.109)  and  the  percentage  of  polymorphic  loci 
(26.3  percent)  of  P.  cembra,  as  compared  to  other  species. 
However,  it  should  be  noted,  that  the  only  sample  of  this  spe- 
cies studied  by  us  represents  a  small  marginal  stand  from 
the  eastern  Carpathians,  while  central  populations  may 
appear  to  be  more  variable.  The  limited  nimaber  of  loci 
studied  in  the  only  work  (Szmidt  1982)  dealing  with  allo- 
zyme polymorphism  in  Swiss  stone  pine  over  the  species 


37 


Table  2 — Levels  of  allozyme  variability  in  stone  pine  populations 


Species  and  Number  Number 


population  names 

of  loci 

of  trees 

P2 

Ho' 

He' 

Dim  1^  ^ihifS^o 
ninus  SiOiriCa 

Filin  Klyuch^  (F«) 

19 

53.0 

1.8 

47.4 

0.155 

0.152 

Mutnaya  Rechka  (M) 

19 

75.0 

1.9 

47.4 

.137 

.140 

Maiyi  Kebezh  (MK) 

19 

15.0 

1.8 

52.6 

.158 

.153 

Sobach'ya  Rechka  (SR) 

19 

12.0 

1.7 

42.1 

.145 

.173 

Listvyanka  (L) 

19 

41.0 

1.9 

42.1 

.163 

.151 

Yayl'u  (YA) 

19 

43.0 

1.9 

47.4 

.153 

.154 

Smokotino  (S) 

19 

43.0 

1.7 

42.1 

.168 

.161 

Zorkal'tsevo-I  (Z-1) 

19 

34.0 

1.7 

47.4 

.198 

.176 

Zorkartsevo-3  (Z-3) 

19 

34.0 

1.7 

47.4 

.184 

.165 

Mean 

19 

38.9 

1.8 

46.2 

.162 

.158 

UsfChornaya  (UCH) 

19 

15.0 

1.5 

26.3 

.128 

.109 

Pinus  pumila 

Grustnyi  (GR) 

20 

54.0 

2.3 

50.0 

.230 

.239 

Uttyveem  (UT) 

20 

55.0 

2.5 

60.0 

.243 

.251 

Kamenskoye  (K) 

20 

55.0 

2.3 

70.0 

.269 

.258 

Mean 

20 

54.7 

2.4 

60.0 

.247 

.249 

Pinus  koraiensis 

Malokhekhtsirskoye  (MKH) 

17 

30.0 

1.6 

47.1 

.147 

.132 

Khor  (KHO) 

17 

16.0 

1.6 

47.1 

.162 

.148 

OlKllOie-Mlin  ^OA; 

1  7 
1  / 

1  1  .u 

1  .O 

.Mo 

Mean 

17 

19.0 

1.6 

43.2 

.137 

.131 

Pinus  albicaulis 

Livingstone  Falls  (LF) 

16 

52.0 

1.9 

56.3 

.213 

.204 

'Mean  number  of  alleles  per  locus. 

^Mean  percent  of  loci  polymorphic  (locus  has  been  considered  polymorphic  if  the  frequency  of  the  most  com- 
mon allele  has  not  exceeded  95  percent). 
^Mean  heterozygosity  observed. 
"Mean  heterozygosity  expected. 
^See  text  for  origins  of  populations. 
^Abbreviated  denomination  of  population. 


range  does  not  allow  any  conclusions  to  be  made  about  the 
variability  level. 

Heterozygosity  of  Korean  pine  (mean  0.131)  is  slightly 
higher.  As  to  Pinus  pumila,  this  species  shows  the  highest 
values  of  H  among  stone  pines  (0.249),  close  to  the  maxi- 
mum ones  known  for  conifers  (Politov  and  others  1992). 
One  cannot  exclude  the  fact  that  such  a  high  variability 
is  very  important  for  adaptation  of  this  species  to  a  wide 
spectnmti  of  environmental  conditions,  including  those  of 
permafrost  regions,  and  drastic  seasonal  climatic  changes. 
Nearly  the  same  heterozygosity  level  (0.204)  was  found  in 
the  only  studied  population  of  whitebark  pine  from  the 
north  moimtain  area. 

Genotype  Distribution  in  Adult  Trees  and 
Embryos — Comparison  of  observed  genotype  distributions 
with  those  expected  from  the  corresponding  Hardy- Weinberg 
proportions  is  commonly  used  in  population  genetics  to  re- 
veal and  identify  various  factors  affecting  genetic  structure 
of  populations.  In  stone  pines  such  analysis  has  not  re- 
vealed pronounced  deviations  from  Hardy- Weinberg  equi- 
librium for  mature  trees  in  any  samples. 


Statistically  significant  excess  or  deficiency  of  heterozy- 
gotes  have  not  been  observed  for  allozyme  loci  studied  in 
stone  pine  populations.  Nevertheless,  the  sign  criterion 
showed  that  there  were  significantly  more  cases  of  a  slight 
heterozygote  excess  (Politov  and  others  1992),  which  may 
mean  existence  of  a  tendency  toward  systematic  excess  of 
heterozygotes  of  alloz3mie  loci  among  adult  trees.  Figure  1 
shows  mean  values  of  Wright's  fixation  index  (or  "inbreed- 
ing coefficient,"  F  =  l-H/Hg,  where  H^  and  Hg  are  mean 
observed  and  expected  heterozygosities  for  all  loci  studied 
in  the  species).  Predominance  of  slightly  negative  or  almost 
zero  F  values  indicates  absence  of  significant  heterozygote 
deficiency  in  the  reproductive  part  of  the  popvilations.  Spe- 
cies mean  values  were:  for  P.  sibirica,  -0.025;  for  P.  cembra, 
-0.174;  for  P.  koraiensis,  -0.046.  F  values  for  P.  pumila 
were  slightly  positive  in  two  samples  and  negative  in  the 
third  one  (mean  F  =  0.008). 

In  most  embryo  samples,  on  the  contrary,  a  slight  defi- 
ciency of  heterozygotes  was  regularly  observed  (Politov  and 
others  1992).  Figure  2  presents  mean  F  values  in  embryos 
and  adult  trees  calculated  for  the  same  loci.  In  embryos 


38 


0.1 
0.05 

F  0 

-0.05 
-0.1 
-0.15 


A. 
'1 

niri 

1 

ij 

1 

Populations:    F    M  MK  SR  L  YA  S  Z-1  Z-3  MEAN 


GR  UT  K      MEAN        MKH      SA  MEAN 


J  KHO 
I  


J 


P.  koraiensis 


P.  sibirica     P.  cembra   P.  pumila 

Figure  1 — Observed  fixation  index  ("inbreeding  co- 
efficient") values  (F)  in  samples  of  adult  trees.  All 
studied  loci  used.  (For  full  names  and  origins  of 
populations  see  text  and  table  2). 


the  values  are  positive  in  nearly  all  samples  (except  one), 
whereas  mature  trees  display  mainly  negative  F  values. 

Probable  Causes  of  Heterozygote  Deficiency — 

most  probable  cause  for  heterozygote  deficiency  at  early  de- 
velopmental stages  of  stone  pines  may  be  inbreeding.  The 
species  studied,  as  well  as  other  conifers,  have  a  mixed  mat- 
ing system  (Brown  and  others  1985;  Politov  and  Krutovskii 
1990  and  references  therein)  characterized  by  a  random 
outcrossing  with  a  certain  percentage  of  selfing.  Heterozy- 
gote deficiency  may  theoretically  be  caused  by  both  self- 
pollination  and  other  consanguineous  matings  (Ritland  and 
El-Kassaby  1985;  Shaw  and  others  1981).  One  cannot  also 
exclude  the  action  of  natural  selection  against  heterozy- 
gotes  at  a  certain  developmental  stage. 

Much  evidence  has  been  reported  demonstrating  hetero- 
zygote deficiency  among  embryos  in  populations  of  conifers 
(summarized  by  Bush  and  Smouse,  1992).  In  some  rare 
cases  deficiency  is  revealed  in  mature  trees.  The  authors 
explained  this  phenomenon  by  inbreeding  and,  particularly, 
by  selfing. 

Mating  System  Analysis — ^The  fact  that  selfing  is  most 
commonly  regarded  as  a  factor  causing  inbreeding  in  conifer 
species  has  good  reasons.  Coincidence,  in  general,  of  polli- 
nation with  pistil  receptivity  of  the  same  tree  and  absence  of 
effective  self-sterility  mechanisms  make  this  process  highly 
probable.  In  order  to  test  the  h3T)othesis  that  partial  selfing 
is  the  main  cause  of  heterozygote  deficiency  in  stone  pines, 
we  have  applied  methods  of  mating  system  parameters  es- 
timation based  on  allozyme  data  and  mixed  mating  system 
theory  (Brown  and  others  1985).  Use  of  isozymes  and  pow- 
erful mathematical  procedures  (Expectation-maximization 
method,  Newton-Raphson  method,  bootstrap,  etc.)  have 
made  it  possible  to  obtain  quantitative  estimates  of  selfing 
and  outcrossing  rates  in  natural  conifer  populations.  For 
different  species  percentage  of  selfing  varies  between  2  to  3 


and  20  to  30  percent  (see  Brown  and  others  1985;  Politov 
and  Krutovskii  1990,  for  review).  For  stone  pines  such 
estimates  have  not  been  made  until  recently.  Here  we 
present  data  obtained  on  three  stone  pine  species. 

The  results  of  mating  system  analysis  of  Siberian  stone 
pine  have  been  published  earlier  (Politov  and  Krutovskii 
1990).  If  we  consider  a  randomly  outcrossing  population 
at  eqmlibrium,  absence  of  pre-  and  postzygotical  selections, 
sexual  symmetry,  separated  generations,  etc.,  the  expected 
inbreeding  coefficient  Fg  can  be  derived  from  the  equation 
Fg  =  (l-^)/(l+^),  where  t  is  the  outcrossing  rate  estimated 
for  this  species.  In  our  study  the  expected  inbreeding  coef- 
ficient Fg  has  turned  out  to  be  very  close  to  the  actually 
observed  one  (F^)  [F„  =  iK~R^)fH^,  where  Hg  and  H^  are, 
correspondingly,  the  expected  from  Hardy- Weinberg  pro- 
portions and  the  directly  observed  heterozygosities]  in  em- 
bryo samples.  Thus,  selfing  (1-t)  whose  estimated  rate  is, 
on  the  average,  as  high  as  15  percent  (from  2  to  27  in  indi- 
vidual populations)  is  most  likely  to  be  one  of  the  main  fac- 
tors of  heterozygote  deficiency  occurrence  among  embryos 
of  this  species. 

Available  data  for  three  stone  pine  species  are  presented 
in  table  3.  Estimation  of  outcrossing  rates  was  made  based 
on  four  to  eight  polymorphic  loci  and  using  two  different 
computation  procedures — variants  of  both  single-  and  mtdti- 
locus  methods  (Ritland  and  El-Kassaby  1985;  Ritland  and 
Jain  1981).  Both  procedures  have  been  realized  by  Ritland 
(1990)  in  his  original  computer  program  MLT.  The  multi- 
locus  estimation  (which  has  more  detective  power  than  the 
single-locus  one)  gives  for  Pinus  sibirica  populations  values 
ranging  from  0.846  to  0.980  (mean  0.894),  and  for  P.  korai- 
ensis values  fi:'om  0.920  to  1.034  (mean  0.974).  The  out- 
crossing rate  foimd  in  P.  cembra  was  substantially  lower 


P.  sibirica 


P.  cembra  i_ 


Q  Adult  trees 
I  Embryos 


Figure  2 — Observed  fixation  index,  "inbreeding 
coefficient"  (F^),  values  in  samples  of  adult  trees 
(striped  bars)  and  embryos  (filled  bars).  Only 
eight  common  loci  used.  Fg  is  expected  inbreed- 
ing coefficient  [Fg  =  {^-tJ^{^+tJ].  S  is  selfing 
rate  (s  =  l-f^,  see  table  3).  (For  full  names  and 
origins  of  populations  see  text  and  table  2). 


P.  lioraiensis 


39 


Table  3 — Multilocus  (f  J  and  single-locus  {Q  estimates  of  outcrossing 
rate  in  stone  pine  populations 


Species  and         Number   Outcrossing  rate 

population  names      of  loci         t„  (s.e.)  (s.e.) 


Pinus  sibirica 

Filin  Klyuch-85^ 

8.0 

0.817 

(0.031) 

0.788 

(0.030) 

Filin  Klyuch-86 

8.0 

.892 

(.042) 

.884 

(.052) 

Mutnaya  Rechka-86 

8.0 

.980 

(.019) 

.959 

(.048) 

iviuiiidyct  nuoiirVd  0/ 

8.0 

.863 

\.\JOO) 

.0^0 

1  n'^fi^ 

Maiyi  Kebezh 

8.0 

.960 

(.072) 

.933 

(.060) 

Sobach'ya  Rechka 

8.0 

.846 

(.125) 

.834 

(.105) 

^lOlVyCII  IrVCl 

8.0 

.912 

(.031) 

fl71 

Yayl'u 

8.0 

.855 

(.045) 

.776 

(.043) 

Smokotino 

7.0 

.929 

(.033) 

.886 

(.037) 

Mean 

7.9 

.894 

(.057) 

.862 

(.054) 

Pinus  cembra 

Ust'-Chornaya 

4.0 

.686 

(.025) 

.707 

(.045) 

Pinus 

koraiensis 

Malokhekhtsirskoye 

6.0 

.967 

(.062) 

.914 

(.052) 

Khor 

7.0 

.920 

(.037) 

.929 

(.066) 

Sikhote-Alin' 

6.0 

1.034 

(.070) 

.964 

(.031) 

Mean 

6.3 

.974 

(.058) 

.936 

(.051) 

'Number  means  year  of  sampling  of  material;  see  text  for  origins  of 
populations. 


(0.686),  but  we  regard  this  value  only  as  a  preliminary  one 
because  of  the  limited  sample  size  and  number  of  loci  stud- 
ied. Nevertheless,  it  may  also  be  the  real  level  of  selfing, 
probably  due  to  the  limited  nimiber  of  trees  in  the  studied 
locality  and  a  low  stand  density. 

Shaw  and  others  (1981)  note  that  we  can  conclude  the  role 
of  inbreeding  components  other  than  selfing  per  se  in  con- 
sanguineous matings  (for  example,  mating  between  closely 
related  neighbor  trees  in  a  population  with  family  structure 
or  assortative/selective  pollination/fertilization  between 
genetically  similar  trees)  from  the  difference  between  the 
values  of  multi-  and  single-locus  estimates.  In  oiu*  study  of 
stone  pines  these  differences,  for  nearly  all  samples,  appesir 
to  be  negligible  (the  difference,  in  general,  does  not  exceed 
0.04).  This  implies  a  rather  weak  effect  of  mating  between 
relatives  or,  in  other  words,  absence  of  microsubdivision  of 
the  population. 

Fumier  and  others  (1987)  found  that  genetic  and  physical 
distances  between  clumps  of  Pinus  albicaulis  fail  to  corre- 
late. The  authors  related  this  fact  to  the  effective  "mixing" 
of  seeds  in  the  process  of  their  dispersal  by  the  Clark's  nut- 
cracker {Nucifraga  columbiana  Engelm).  We  can  expect  to 
observe  randomization  of  such  kind  (more  closely  spaced 
seedlings  or  clumps  are  not  more  related  ones)  in  any  other 
pine  species  depending  on  the  European  nutcracker  (AT.  cary- 
ocatactes  L.)  in  their  dispersal,  since  the  effectiveness  of  nut- 
crackers as  seed  dispersers  has  been  repeatedly  proved  by 
numerous  investigations  (Lanner  1980,  1990;  Tomback 
and  Linhart  1991). 

Selection  Against  Inbred  Progeny — What  may  cause 
"disappearance"  of  heterozygote  deficiency  observed  among 
embryos,  as  compared  to  adult  trees  of  the  population 


(figs.  1  and  2)?  This  can,  most  likely,  be  the  result  of  selec- 
tion against  inbred  progeny  formed  by  self-pollination,  as 
well  as  selection  in  favor  of  heterozygotes  (overdominance). 
Both  selections  may  increase  heterozygosity  up  to  the  equi- 
librium level,  but  only  as  a  result  of  negative  assortative 
fertilization  or  selection  in  favor  of  heterozygotes  (balancing 
selection)  can  heterozygosity  also  exceed  Hardy-Weinberg 
expectations.  This  process  can  be  detected  with  the  aid  of 
isozyme  gene  markers  if  they  are  either  involved  in  it  di- 
rectly or  £ire  in  gametic  linkage  disequilibrium  with  selec- 
tively significant  genes  (Crow  and  Kimura  1970). 

In  naturgJ  conifer  populations  heterozygosity  of  adult 
trees  is  often  higher  than  in  embryos  (see  Bush  and  Smouse 
1992;  Krutovskii  and  others  1988;  Politov  and  others  1992, 
for  review).  For  instance,  in  Pinus  radiata  (Plessas  and 
Strauss  1986)  heterozygosity  of  yoimg  trees  (17  to  20  years) 
exceeded  that  of  5-year  seedlings  and  in  the  latter  it  was 
higher  than  in  embryos.  Yazdani  and  others  (1985)  found 
higher  heterozygosity  of  young  £ind  adult  Scots  pine  trees 
as  compared  to  embryos;  differences  between  the  young  and 
the  adult  trees  were  slight  indicating  the  action  of  selec- 
tion exactly  at  early  stages.  For  stone  pines  we  can  suggest 
a  pEirticular  life  stage  of  natural  selection  acting  against 
homozygotes — ^when  groups  of  seedlings  begin  to  germinate 
out  of  nutcrackers'  caches  competing  for  resources.  It  is 
noteworthy  that  Eurasian  stone  pines  (as  compared  to  white- 
bark  pine)  display  multistem  forms  rarely,  which  may  be 
the  result  of  intensive  competition  between  seedlings  of  the 
same  cache. 

More  direct  evidence  was  obt£iined  in  experiments  carried 
out  by  Farris  and  Mitton  (1984),  who  studied  embryos  and 
seedlings  of  yellow  pine  {Pinus  ponderosa)  and  foimd  the 
increase  of  heterozygosity  with  age.  Muona  and  others 
(1987)  reveeded  heterozygosity  deficit  in  Scots  pine  embryos 
and  its  "disappearance"  in  1.5-year-old  seedlings. 

Inbred  progeny  elimination  at  early  developmental  stages 
in  conifers  is  also  supported  by  some  nonallozyme  data  (see 
for  example,  Koski  1973). 

CONCLUSIONS 

Study  of  alloz5rme  variability  in  stone  pines  has  shown 
the  genetic  diversity  tjrpical  to  widespread  and  abimdant 
conifer  species  with  continuous  ranges.  Differences  between 
stone  pines  can  be  explained  by  both  the  events  of  their 
recent  evolutionary  history  and  adaptive  significance  of 
the  allozyme  variation.  Heterozygote  deficiency  among  em- 
bryos and  slight  heterozygote  excess  in  mature  trees  are 
common  features  of  stone  pines,  as  well  as  other  conifers. 
A  substantial  part  of  self-pollinated  progeny  is  eUminated 
during  ontogenesis  by  natural  selection. 

ACKNOWLEDGMENTS 

We  are  greatly  thankful  to  G.  R.  Furnier,  V.  N.  Vorobjev, 
N.  A.  Vorobjeva,  L.  I.  Milyutin,  R.  T.  Gut,  and  P.  Raztmiov 
for  providing  seeds  and  aid  in  their  sampling,  to  Yu.  S. 
Belokon'  for  helpful  assistance  in  the  laboratory  analysis, 
to  Yu.  P.  Altukhov  for  constant  support  of  our  study  and 
very  useful  conmients  on  this  meinuscript,  and  to  K.  Ritland 
for  kindly  providing  us  with  the  computer  program  MLT. 


40 


We  also  appreciate  H.-R.  Gregorius's  invaluable  comments 
that  helped  to  substantially  improve  the  manuscript. 

REFERENCES 

Adams,  W.  T.;  Strauss,  S.  H.;  Copes,  D.  L.;  Griffin,  A.  R., 
eds.  1992.  Population  genetics  of  forest  trees:  Proceed- 
ings of  the  International  Symposiimi  on  Population  Ge- 
netics of  Forest  Trees;  1990  July  31-August  2;  Corvallis, 
OR,  U.S.A.  Kluwer  Academic  Publishers.  420  p.  (see  also 
New  Forests.  1992.  6.) 

Bobrov,  E.  G.  1978.  Forest-forming  conifers  of  the  USSR. 
Moscow:  Nauka.  190  p.  [In  Russian]. 

Brown,  A.  H.  D.;  Barrett,  S.  C.  H.;  Moran,  G.  F.  1985.  Mat- 
ing system  estimation  in  forest  trees:  Models,  methods 
and  meanings.  In:  Gregorius,  H.-R.,  ed.  Population  genet- 
ics in  forestry.  Lecture  Notes  in  Biomathematics.  Vol.  60. 
Berlin-Heidelberg-New  York-Tokyo:  Springer- Verlag: 
32-49. 

Bush,  R.  M.;  Smouse,  P.  E.  1992.  Evidence  for  the  adaptive 
significance  of  allozymes  in  forest  trees.  New  Forests.  6: 
179-196. 

Critchfield,  W.  B.;  Little,  E.  L.  1966.  Geographic  distribution 
of  the  pines  of  the  world.  Misc.  Publ.  991.  Washington, 
DC:  U.S.  Department  of  Agriculture.  99  p. 

Crow,  J.  F.;  Kimura,  M.  1970.  An  introduction  to  popula- 
tion genetics  theory.  New  York:  Harper  and  Row. 

Farris,  M.  A.;  Mitton,  J.  B.  1984.  Population  density,  out- 
crossing rate  and  heterozygote  superiority  in  ponderosa 
pine.  Evolution.  38(5):  1151-1154. 

Fineschi,  S.;  Cannata,  F.;  Malvolti,  M.  E.;  Hattemer,  H.  H., 
eds.  1991.  Biochemical  markers  in  the  population  genet- 
ics of  forest  trees:  Proceedings  of  the  Joint  Meeting  of  the 
Working  Parties  S2. 04-021  Population  Genetics  and  Eco- 
logical Genetics  and  S2. 04-05  Biochemical  Genetics  of 
the  International  Union  of  Forest  Research  Organizations 
(lUFRO);  October  1988;  Porano,  Italy.  SPB  The  Hague: 
Academic  PubHshing  bv.  251  p. 

Fumier,  G.  F.;  Knowles,  P.;  Aleksiuk,  M.  A.;  Dancik,  B.  P. 

1986.  Inheritance  and  linkage  of  allozymes  in  seed  tissue 
of  whitebark  pine.  Canadian  Journal  of  Genetics  and 
Cytology.  28(4):  601-604. 

Fumier,  G.  R.;  Knowles,  P.;  Clyde,  M.  A.;  Dancik,  B.  P. 

1987.  Effect  of  avian  seed  dispersal  on  the  genetic  struc- 
ture of  whitebark  pine  popidations.  Evolution.  41(3): 
607-612. 

Goncharenko,  G.  G.;  Padutov,  V.  E.;  Krutovskii,  K  V.; 
Podzharova,  Z.  S.;  Kirgizov,  N.  Ya.;  Politov,  D.  V.  1988. 
Level  of  genetic  variability  in  Pinus  sibirica  in  Altai. 
Doklady  Akademii  Nauk  SSSR.  299(1):  222-225.  [In 
Russian].  [English  translation  published  by  Pleniim 
PubHshing  Corp.  in  Doklady  Biological  Sciences;  1988 
September:  139-141]. 

Hamrick,  J.  L.;  Godt,  M.  J.  1989.  Allozyme  diversity  in 
plant  species.  In:  Brown,  A.  H.  D.;  Clegg,  M.  T.;  Kahler, 
A.  L.;  Weir,  B.  S.,  eds.  Plant  population  genetics  breeding 
and  genetic  resources.  Sanderland,  MA:  Sinauer  Associ- 
ates: 43-63. 

Koski,  V.  1973.  On  self-fertilization,  genetic  load  and  sub- 
sequent inbreeding  in  some  conifers.  Comm.  Inst.  For, 
Fenn.  78(10):  1-40. 


Krutovskii,  K.  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.  1987. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  I.  (jenetic  control  of  isozyme  systems.  Genetika. 
23(12):  2216-2228.  [In  Russian].  [EngHsh  translation 
published  by  Plenimi  Publishing  Corp.  in  Soviet  Genet- 
ics; 1988  June:  1568-1576]. 

Krutovskii,  K  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.  1988. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  II.  Level  of  allozyme  variability  in  Western 
Sayan.  Genetika.  24(1):  118-124.  [In  Russian].  [EngHsh 
translation  published  by  Plenum  Publishing  Corp.  in 
Soviet  Genetics;  1988  July:  81-86]. 

Krutovskii,  K  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.;  Milyutin, 
L.  I.;  Kxiznetsova,  G.  V.;  Iroshnikov,  A.  I.;  Vorob'ev,  V.  N.; 
Vorob'eva,  N.  A.  1989.  Genetic  variability  in  Siberian 
stone  pine,  Pinus  sibirica  Du  Tour.  IV.  Genetic  diversity 
and  differentiation  between  populations,  (jenetika.  25(11): 
2009-2032.  [In  Russian].  [EngHsh  translation  pubHshed 
by  Plenimi  Publishing  Corp.  in  Soviet  Cjenetics;  1990  May: 
1343-1362]. 

Krutovskn,  K.  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.  1990.  In- 
terspecific genetic  differentiation  of  Eurasian  stone  pines 
for  isoenzyme  loci.  Genetika.  26(4):  694-707.  [In  Russian]. 
[English  translation  published  by  Plenum  Publishing 
Corp.  in  Soviet  Genetics;  1990  October:  440-450]. 

Krutovskii,  K  V.;  Politov,  D.  V.  1992.  Study  of  intra-  and 
interspecific  genetic  differentiation  between  Eurasian 
cedar  pines  using  isozyme  loci  and  multidimensional 
analysis.  In:  Symposium  on  molecular  mechanisms  of 
genetic  processes:  proceedings.  Moscow:  Nauka  Publish- 
ers: 87-96.  [In  Russian]. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in  the 
ecology  and  evolution  of  limber  and  whitebark  pines.  In: 
6th  North  American  Forest  Biology  Workshop:  proceed- 
ings; 1980  August  11-13.  Edmonton,  AB:  University  of 
Alberta:  15-47. 

Lanner,  R.  M.  1990.  Biology,  taxonomy,  evolution,  and 
geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposimn  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource.  Gen.  Tech. 
Rep.  INT-270.  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermoimtain  Research  Station:  14-24. 

Ledig,  F.  T.  1986.  Heterozygosity,  heterosis,  and  fitness 
in  outbreeding  plants.  In:  Soule,  M.  E.,  ed.  Conservation 
biology:  the  science  of  scarcity  and  diversity:  77-104. 

Mvdler-Starck,  G.;  Ziehe,  M.,  eds.  1991.  Genetic  variation 
in  Eiiropean  populations  of  forest  trees:  proceedings  of 
the  workshop  on  genetic  variation  in  forest  tree  popula- 
tions in  Europe;  1990;  Gottingen,  Germany.  Frankfurt 
am  Main:  Sauerlander's  Verlag.  271  p. 

Muona,  O.;  Yazdani,  R.;  Rudin,  D.  1987.  Genetic  change 
between  life  stages  in  Pinus  syluestris:  allozjone  varia- 
tion in  seeds  and  planted  seedlings.  Silvae  Genetica. 
36(1):  39-42. 

Plessas,  M  E.;  Strauss,  S.  H.  1986.  Allozyme  differentiation 
among  populations,  stands,  and  cohorts  in  Monterey  Pine. 
Canadian  Journal  of  Forest  Research.  16(6):  1155-1164. 

Podogas,  A.  v.;  Shurkhal,  A.  V.;  Semerikov,  V.  L.;  Rakitskaya, 
T.  A.  1991.  Genetic  variabiHty  of  needle  tissue  enzymes 
of  Siberian  cedar  pine  (Pinus  sibirica  Du  Tour).  Genetika. 


41 


27(4):  695-703.  [In  Russian].  [English  translation  pub- 
lished by  Plenum  Publishing  Corp.  in  Soviet  Genetics; 
1991  October]. 
Politov,  D.  V.  1989.  Allozyme  polymorphism,  genetic  dif- 
ferentiation, and  mating  system  in  Siberian  stone  pine 
Pinus  sibirica  Du  Tour.  IVIoscow:  Institute  of  General 
Genetics,  Russian  Academy  of  Sciences.  190  p.  Thesis. 
[In  Russian]. 

Pohtov,  D.  v.;  Krutovskii,  K.  V.;  Altukhov,  Yu.  P.  1989. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  III.  Linkage  relationships  among  isozyme  loci. 
Genetika.  25(9):  1606-1618.  [In  Russian].  [EngHsh  trans- 
lation published  by  Plenum  Publishing  Corp.  in  Soviet 
Genetics;  1990  IVIarch:  1053-1063]. 

PoHtov,  D.  v.;  Krutovskii,  K.  V.  1990.  Genetic  variability 
in  Siberian  stone  pine,  Pinus  sibirica  Du  Toiu*.  V.  Analy- 
sis of  mating  system.  Genetika.  26(11):  2002-2011.  [In 
Russian].  [English  translation  published  by  Plenmn  Pub- 
hshing  Corp.  in  Soviet  Genetics;  1991  May:  1309-1316]. 

Pohtov,  D.  v.;  Krutovskii,  K.  V.;  Altukhov,  Yu.  P.  1992. 
Isozyme  loci  characteristics  of  gene  banks  of  populations 
of  cedar  pines.  Genetika.  28(1):  93-114.  [In  Russian]. 
[English  translation  published  by  Plenvun  Publishing 
Corp.  in  Soviet  Genetics;  1992  July:  76-95]. 

Ritland,  K.  1990.  A  series  of  FORTRAN  computer  programs 
for  estimating  plant  mating  system.  Journal  of  Heredity. 
81:  235-237. 


Ritland,  K.;  El-Kassaby,  Y.  A.  1985.  The  nature  of  inbreed- 
ing in  a  seed  orchard  of  Douglas  fir  as  shown  by  an  effi- 
cient multilocus  model.  Theoretical  and  Applied  Genetics. 
71:  375-384. 

Ritland,  K.;  Jain,  S.  1981.  A  model  for  the  estimation  of 
outcrossing  rate  and  gene  frequencies  using  n  indepen- 
dent loci.  Heredity.  47(1):  35-52. 

Shaw,  D.  v.;  Kahler,  A.  L.;  Allard,  R.  W.  1981.  A  multilocus 
estimator  of  mating  system  parameters  in  plant  popula- 
tions. Proceedings  of  the  National  Academy  of  Sciences 
(USA).  78:  1298-1302. 

Strauss,  S.  H.;  Bosquet,  J.;  Hipkins,  V.  D.;  Hong,  Y.-P.  1992. 
Biochemical  and  molecular  genetic  markers  in  biosys- 
tematic  studies  of  forest  trees.  New  Forests.  6:  125-158. 

Szmidt,  A.  E.  1982.  Genetic  variation  in  isolated  popula- 
tions of  stone  pine  {Pinus  cembra).  Silva  Fennica.  16(2): 
196-200. 

Szmidt,  A.  E.;  IMuona,  O.  1989.  Linkage  relationships  of 
allozyme  loci  in  Pinus  sylvestris.  Hereditas.  Ill:  91-97. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1991.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Biology.  4:  185-219. 

Yazdani,  R.;  Lindgren,  D.;  Rudin,  D.  1985.  Gene  dispersion 
and  selfing  frequency  in  a  seed-tree  stand  of  Pinus  sylves- 
tris (L.).  In:  Gregorius,  H.  R.,  ed.  Population  genetics  in 
forestry.  Lecture  Notes  in  Biomathematics.  Vol.  60.  Berlin- 
Heidelberg-New  York-Tokyo:  Springer- Verlag:  139-154. 


42 


GENETIC  POPULATION  STRUCTURE 
AND  GROWTH  FORM  DISTRIBUTION 
IN  BIRD-DISPERSED  PINES 

Diana  F.  Tomback 
WiUiam  S.  F.  Schuster 


Abstract — ^At  least  nine  species  of  the  20  wingless-seed  pines 
are  dependent  on  the  nutcrackers  {Nucifraga)  for  seed  dispersal. 
Seed  dispersal  by  nutcrackers  influences  several  aspects  of  popu- 
lation structure:  (1)  tree  dispersion  patterns  and  growth  form  dis- 
tribution, (2)  genetic  relationships  among  trees  within  popula- 
tions, and  (3)  genetic  diversity  among  populations.  Studies  to 
date  of  three  species,  whitebark  pine  {Pinus  albicaulis),  limber 
pine  {Pinus  flexilis),  and  Swiss  stone  pine  (Pinus  cembra),  pro- 
vide examples  of  intrapopulation  genetic  relationships  and  growth 
form  distribution;  other  recent  work  provides  some  information 
for  interpopulation  comparisons. 


of  parent  trees,  resulting  in  a  haphazard  distribution  of 
seeds  with  respect  to  seed  source  (Tomback  1978,  1982). 
Also,  nutcrackers  often  bury  more  than  one  seed  per  cache 
(mean  of  three  or  fotir  seeds  per  cache  for  Clark's  nut- 
cracker, Tomback  and  Linhart  1990),  and  the  seeds  may 
come  from  the  same  parent  tree  (Tomback  1988;  Tomback 
and  Knowles  1989). 

In  contrast,  most  wind-dispersed  seeds  fall  within  about 
120  m  of  parent  trees  and  are  dependent  on  prevailing 
winds  for  dispersal  (see,  for  example,  McCaughey  and  oth- 
ers 1986).  Updrafts  and  wind  patterns  associated  with 
storms  may  send  wind-dependent  seeds  over  longer  dis- 
tances and  in  other  directions,  and  seed-storing  rodents 
may  move  seeds  aroimd  locally  (Vander  Wall  1992).  How- 
ever, because  the  seeds  of  these  species  usually  move  rela- 
tively short  distances  and  establishment  is  often  in  small, 
disturbed  patches,  a  family  structure  (local  aggregations 
of  related  genotypes)  frequently  results  within  popula- 
tions (see,  for  example,  Knowles  1984;  Linhart  1989; 
Linhart  and  others  1981;  see  also  Fumier  and  others  1987; 
Tomback  and  Linhart  1990). 

These  different  seed-dispersal  modes,  bird  vs.  wind, 
may  lead  to  different  genetic  population  structures.  Ge- 
netic population  structure  resiilts  from  nonrandom  spatial 
distribution  of  genotypes  (Epperson  1990,  and  references 
therein).  Plant  populations  may  show  structure  on  differ- 
ent scales,  such  as  among  populations  and  within  popula- 
tions. The  common  methods  for  investigating  genetic 
population  structure  are  based  on  allozyme  analysis, 
comparing  populations  or  subpopxilations  by  means  of  ei- 
ther allele  frequencies,  Wright's      or  Nei's  Gg.^,  statistics, 
or  spatial  autocorrelation. 

Seed  dispersal  by  nutcrackers  influences  several  levels 
of  popiilation  structure  in  wingless-seed  pines.  Within 
populations,  it  affects  both  (1)  tree  dispersion  patterns 
(whether  individuals  are  clumped  vs.  spaced)  and,  conse- 
quently, growth  form  distribution  and  (2)  genetic  relation- 
ships among  trees.  Among  populations  it  affects  genetic 
diversity.  In  this  paper,  we  explore  the  specific  impacts 
of  nutcracker  seed-dispersal  behavior  on  each  of  these 
aspects  of  population  structure. 

GROWTH  FORM  DISTRffiUTION 


Of  the  20  species  of  pines  with  large,  wingless  seeds 
(collectively  considered  "stone  pines"),  nine  are  known 
dependents  on  the  nutcrackers,  Nucifraga  (Corvidae),  for 
seed  dispersal  (Tomback  and  Linhart  1990  and  references 
therein).  Five  of  these  bird-pines  are  in  the  Subsection 
Cembrae,  two  in  Strobi,  and  two  in  Parry  a;  for  classifica- 
tion see  Critchfield  and  Little  (1966).  For  several  species 
of  these  bird-dependent  pines,  other  corvids,  squirrels,  or 
other  rodents  may  also  make  varying  contributions  to  tree 
estabHshment  (see,  for  example,  Hayashida  1989;  Ligon 
1978;  Vander  Wall  1992),  but  the  factor  in  common  to  all 
dispersers  is  the  tendency  to  remove  seeds  from  the  vicin- 
ity of  parent  trees  and  "scatterhoard"  them  (biuy  small 
clusters  of  seeds  in  subterranean  caches  for  future  use). 
Clark's  nutcrackers  (A^.  columbiana)  bury  one  to  15  or 
more  seeds  per  cache  at  distances  of  several  meters  to 
22  km  from  parent  trees  (Hutchins  and  Lanner  1982; 
Tomback  1978,  1982;  Vander  Wall  and  Balda  1977).  Eur- 
asian nutcrackers  (A^.  caryocatactes)  bury  up  to  24  seeds 
per  cache  and  travel  similar  distances  (see,  for  example, 
Bibikov  1948;  Mattes  1982;  Mezhenny  1961;  Turcek 
1966). 

Thus,  nutcrackers  create  a  highly  impredictable  seed 
shadow,  with  seedhng  establishment  oft^n  occurring  sev- 
eral kilometers  from  parent  trees  (Tomback  and  others 
1990).  Nutcracker-dispersed  seeds  may  end  up  in  small, 
isolated  areas,  particularly  following  disturbances  such 
as  fires.  In  large,  open  or  disturbed  areas,  different  nut- 
crackers may  cache  seeds  from  several  difi"erent  stands 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Diana  F.  Tomback  is  Associate  Professor  of  Ecology,  Department  of 
Biology  and  Center  for  Environmental  Sciences,  University  of  Colorado 
at  Denver,  Denver,  CO,  80217-3364.  William  S.  F.  Schuster  is  Forest  Di- 
rector, Black  Rock  Forest  Consortium,  Box  483,  Cornwall,  NY  12518. 


Research  to  date  on  three  species  of  bird-dispersed 
pines — whitebark  pine  (Pinus  albicaulis)  (Fumier  and 
others  1987;  Linhart  and  Tomback  1985),  limber  pine 
(P.  flexilis)  (Carsey  and  Tomback  1992;  Linhart  and 
Tomback  1985;  Schuster  and  Mitton  1991),  and  Swiss 
stone  pine  (P.  cembra)  (Tomback  and  others  1993) — 


43 


indicates  that  trees  within  populations  occur  in  three 
different  growth  forms.  The  single-trunk  growth  form 
is  typical  of  conifers  and  consists  of  a  single  genet  (indi- 
vidual). Also  composed  of  a  single  genet,  the  multi-trunk 
growth  form  has  two  or  more  trunks  either  contiguous 
at  the  base  or  fused  at  the  base  or  part  way  up  the  bole. 
Multi-trunk  trees  may  be  the  consequence  of  a  tendency 
toward  side-branching  and  lack  of  apical  dominance,  par- 
ticularly under  stressful  environmental  conditions,  such 
as  unstable  soils,  aridity,  or  mechanical  damage  to  the 
leader  (Schuster  and  Mitton  1991;  Weaver  and  Jacobs 
1990).  Although  the  tree  cluster  growth  form  is  also 
composed  of  two  or  more  contiguous  or  fused  trunks,  the 
trunks  consist  of  different  genets  (individuals).  This 
growth  form  results  from  the  germination  and  survival 
of  more  than  one  seedling  from  a  seed  cache  (growth  form 
terminology  from  Tomback  and  others  1990). 

The  latter  two  growth  forms  have  similar  morphologies 
and  thus  cannot  be  distinguished  with  certainty  without 
genetic  analysis  (Schuster  and  Mitton  1991;  Tomback  and 
others  1993)  (figs.  1,  2).  Consequently,  in  field  surveys  of 
frequency  distribution  of  different  growth  forms,  we  com- 
bined the  multi-trunk  and  tree  cluster  forms  together  in 


Figure  1 — Swiss  stone  pine  {Pinus  cembra)  tree 
cluster.  Protein  electrophoresis  indicated  that  each 
trunk  is  a  different  genet. 


Figure  2 — Swiss  stone  pine  {Pinus  cembra) 
multi-trunk  tree.  Protein  electrophoresis  indi- 
cated that  both  trunks  had  the  same  genotype. 


the  category  "tree  clumps"  (term  from  Lanner  1980;  the 
same  as  "multi-trunk  trees  of  unknown  origin,"  Tomback 
and  others  1990).  For  reporting  the  results  of  field  sur- 
veys of  tree  clump  frequency  in  populations,  we  use  the 
term  "tree  site"  to  refer  to  the  small  area  of  ground  sup- 
porting either  a  single-trunk  tree,  a  multi-trimk  tree,  or 
a  tree  cluster. 

Overall,  tree  clumps  accounted  for  19  to  90  percent  of 
the  tree  sites  surveyed  in  three  species  of  bird-dispersed 
pines  (whitebark,  limber,  and  Swiss  stone)  (table  1). 
Whitebark  pine  in  the  two  stands  surveyed  had  the  high- 
est percentage  of  tree  sites  occupied  by  clumps.  In  the 
two  studies  of  limber  pine  combined,  the  percentage  of 
tree  sites  with  clvmips  ranged  from  24  to  48;  the  latter 
population  occurred  on  a  steep,  southeast-facing  slope 
with  gravelly  soil  at  3,310  m  elevation.  Carsey  and 
Tomback  (1992)  also  examined  17  tree  sites  in  a  stand 
of  krummholz  limber  pine  above  timberline  (3,460  m  eleva- 
tion) and  found  clumps  at  all  tree  sites.  For  Swiss  stone 
pine,  clumps  were  found  at  19  to  30  percent  of  the  tree 
sites  and  up  to  45  percent  of  seedling  sites.  In  the  Piz  da 
Staz  study  area,  the  highest  percentage  of  Swiss  stone 
pine  clumps  occurred  at  higher  elevations,  but  this  did 
not  hold  for  the  Muottas  Muragl  study  area. 


44 


Table  1 — Percentage  of  tree  clumps  (combined  multi-trunk  trees  and  tree  clusters)  in  surveyed  populations  of  three  species 
of  bird-dispersed  pines 


Species 

Maximum 

Percent 

and 

Tree 

trunks  per 

sites 

location 

Elevation 

sites 

clump 

with  clumps 

References 

Meters 

Swiss  stone  pine 

Switzerland 

1,835 

406 

6 

21 

Tomback  and 

2,070 

351 

6 

30 

others  1 993 

1 ,835 

96 

4 

29 

2,125 

43 

5 

19 

Whitebark  pine 

California 

2,850 

50 

5+ 

90 

Tomback  and 

Montana 

2,730 

50 

5+ 

58 

Linhart  1990 

Limber  pine 

Colorado 

1,650 

361 

7 

30 

Schuster  and 

Mitton  1991 

Colorado 

2,585 

105 

3 

35 

Carsey  and 

2,810 

135 

4 

24 

Tomback  1992 

3,310 

140 

4 

48 

The  preferred  laboratory  method  for  distinguishing  tree  stone  pine  clumps  sampled  in  one  stand,  70  percent  were 
clusters  from  multi-tnmk  trees  has  been  starch  gel  pro-  tree  clusters  (Tomback  and  others  1993).  Remarkably, 

tein  electrophoresis  to  assay  the  genotype  of  seed  (Fiimier  an  examination  of  the  number  of  gene  loci  used  for  each 
and  others  1987)  or  foliage  tissue  obtained  from  each  study  in  table  2  shows  no  relationship  with  the  percent- 

tnmk  of  a  tree  clump  (Carsey  and  Tomback  1992;  Linhart         age  of  clusters  determined  for  each  sample  (Spearman 
and  Tomback  1985;  Schuster  and  Mitton  1991;  Tomback  rank  correlation,    =  -0.406,  n  =  12,  NS).  With  the  excep- 

and  others  1993;  for  review  of  technique,  see  Kephart  tion  of  the  Umber  pine  clumps  analyzed  by  Schuster  and 

1990).  Those  clvunps  that  have  two  or  more  trimks  with  Mitton  (1991),  which  represented  all  the  climips  in  the 

different  genotypes  are  classified  as  tree  clusters.  Unfor-  popialation,  the  clvmips  examined  in  the  studies  in  table  2 

timately,  this  technique  may  underestimate  slightly  the 
nimibers  of  tree  clusters,  because  seeds  within  nutcracker 
caches  tend  to  be  genetic  relatives  and  thus  may  inherit 
similar  genotypes  (Carsey  and  Tomback  1992;  Schuster 
and  Mitton  1991).  In  a  few  cases  tree  cluster  members 
may  possess  identical  genotypes  at  the  gene  loci  exam- 
ined, so  the  clusters  are  misclassified  as  multi-trunk 
(single-genotype)  trees.  The  proced\ires  for  calculating 
the  probability  of  such  errors  are  presented  in  Schuster 
and  Mitton  (1991). 

For  studies  of  whitebark  pine,  hmber  pine,  and  Swiss 
stone  pine  combined,  genetic  analysis  indicated  that  18 
to  100  percent  of  the  tree  clumps  sampled  were  tree  clus- 
ters (table  2).  The  low  value,  18  percent,  was  obtained  by 
Schuster  and  Mitton  (1991)  for  a  limber  pine  population 
in  the  Pawnee  National  Grasslands,  at  the  periphery  of 
the  range  of  both  the  pine  and  Clark's  nutcracker.  The 
high  values  of  100  percent  for  limber  pine  and  83  percent 
for  whitebark  pine  were  based  on  small  sample  sizes 
in  =  six  clumps  in  both  cases,  Linhart  and  Tomback  1985) 
Filmier  and  others  (1987)  found  that  58  and  70  percent, 
respectively,  of  the  whitebark  pine  clumps  they  sampled 
in  two  stands  contained  more  than  one  genotype;  they  did 
not  sample  all  trunks  of  some  clumps,  so  these  percent- 
ages are  conservative.  Clumps  in  four  limber  pine  stands 
were  studied  by  Carsey  and  Tomback  (1992),  with  per- 
centages of  clusters  ranging  from  24  percent  above  tim- 
berhne  to  81  percent  at  2,810  m  elevation.  Of  the  Swiss 


Table  2 — Percentage  of  tree  clumps  (combined  multi-trunk  trees 
and  tree  clusters)  found  to  be  tree  clusters  (growth  form 
with  more  than  one  genet)  in  three  species  of  bird- 
dispersed  pines 


Species 

Clumps 
examined 

Gene 
loci 
used 

Percent 
clusters 

References 

Swiss  stone 

23 

4 

70 

Tomback  and  others 

pine 

1993 

Whitebark  pine 

6 

4 

83 

Linhart  and  Tomback 

1985 

12 

11 

58 

Furnier  and  others 

23 

11 

70 

1987 

Limber  pine 

7 

2 

57 

Linhart  and  Tomback 

6 

4 

83 

1985 

6 

4 

100 

108 

10 

18 

Schuster  and  Mitton 

1991 

18 

7-9 

56 

Carsey  and  Tomback 

21 

7-9 

81 

1992 

18 

7-9 

44 

17 

7-9 

24 

45 


represented  a  sample  of  the  total  number  of  climaps  present 
and  were  sometimes  not  selected  entirely  at  random.  For 
the  studies  by  Furnier  and  others  (1987)  and  Carsey  and 
Tomback  (1992),  only  reproductively  mature  tree  clumps 
were  sampled;  for  the  studies  by  Linhart  and  Tomback 
(1985)  and  Tomback  and  others  (1993),  clumps  were  usu- 
ally sampled  as  encountered,  but  an  effort  was  made  to 
include  some  seedlings  and  young  trees. 

Only  three  studies  permit  an  estimate  of  the  overall 
percentage  occurrence  of  tree  clusters  among  tree  sites 
in  populations  (table  3).  Limber  pine  tree  clusters  in  the 
Pawnee  National  Grasslands  population  occurred  at  only 
5  percent  of  the  total  tree  sites.  For  limber  pine  in  the 
Front  Range,  clusters  occurred  at  similar  percentages  of 
tree  sites,  19  to  21,  despite  differences  in  elevation.  Clus- 
ters of  Swiss  stone  pine  occupied  21  percent  of  the  tree 
sites  in  the  upper  subalpine  population. 

Apparently,  tree  clusters  typically  occupy  a  substantial 
proportion  of  the  tree  sites  in  each  population  of  whitebark, 
limber,  and  Swiss  stone  pine.  Moreover,  the  percentage 
of  individual  genets  in  a  population  that  are  found  in  clus- 
ters is  much  higher  than  the  proportion  of  tree  sites  sup- 
porting clusters.  Based  on  the  previous  studies,  12  to  40 
percent  of  the  individuals  in  populations  of  bird-dispersed 
pines  occur  in  a  highly  clumped  spatial  pattern,  primarily 
as  a  consequence  of  seed  dispersal  by  nutcrackers. 

If  stress  plays  a  role  in  the  growth  of  the  multi-trunk 
form  from  a  single  genet,  as  proposed  by  Schuster  and 
Mitton  (1991)  and  Weaver  and  Jacobs  (1990),  then  we 
might  predict  a  relatively  higher  percentage  of  multi- 
trunk  trees  in  more  stressful  environments.  In  fact,  about 
25  percent  of  the  limber  pine  tree  sites  in  the  Pawnee 
National  Grasslands  population  support  multi-trunk  trees 
(from  tables  1  and  2).  This  area  is  at  lower  elevations 
(1,650  m)  on  the  plains  east  of  the  Rocky  Mountains,  at 
the  eastern  boundary  of  the  limber  pine  range,  and  trees 
may  be  subjected  to  stress  factors  such  as  grazing,  fire, 
unstable  slopes,  heat,  and  extreme  aridity  (Lepper  1974; 
McNaughton  1984;  Schuster  and  Mitton  1991).  The  high- 
est estimated  percentages  of  multi-trunk  limber  pine  trees 
in  the  Front  Range  populations  studied  by  Carsey  and 
Tomback  (1992)  occurred  in  the  upper  subalpine  popula- 
tion (27  percent,  elevation  3,310  m)  and  in  the  population 
above  timberline  (76  percent,  elevation  3,460  m),  where 
winds  and  accompanying  desiccation  are  extreme  and  the 


Table  3 — Estimated  percentage  of  tree  sites  tliat  support  clusters  of 
individuals  in  populations  of  two  species  of  bird-dispersed 
pines 


Table  4 — Estimated  mean  and  standard  error  of  relatedness  of 
individuals  within  and  between  limber  pine  tree  clusters 


Percent 

Species       clusters  Elevation 


References 


Meters 

Swiss  stone  pine    21    2,050  to  2,240    Tomback  and  others  1 993 


Limber  pine 


5 
19 
19 
21 


1,650 
2,585 
2,810 
3,310 


Schuster  and  Mitton  1 991 
Carsey  and  Tomback  1 992 


Mean    Standard  error 


References 


Within  tree  clusters 

0.19  0.10  Schuster  and  Mitton  1 991 

.43  .13  Carsey  and  Tomback  1 992 

Between  tree  clusters 

.01  .04  Carsey  and  Tomback  1 992 


soil  is  gravelly  and  unstable.  We  also  calculated  that  only 
9  percent  of  the  tree  sites  of  the  upper  subalpine  popula- 
tion of  Swiss  stone  pine  supported  multi-trunk  trees.  In 
contrast  to  the  former  study  areas,  the  Swiss  stone  pine 
tree  sites  are  not  typically  characterized  by  aridity  and 
poor  soils,  which  may  explain  the  relatively  low  frequency 
of  multi-trunk  trees  (Tomback  1988). 

We  should  note  here  that  tree  clusters  are  not  known 
to  occur  in  Colorado  pinyon  pine  (Pinus  edulis),  which  is 
dispersed  by  both  Clark's  nutcrackers  and  pinyon  jays 
(Gymnorhinus  cyanocephalus).  Vander  Wall  and  Balda 
(1977)  report  that  only  one  seed  from  each  cache  grows 
to  maturity. 

GENETIC  RELATIONSHIPS 

Relationships  Within  and  Between  Tree 
Clusters — Because  nutcrackers  tend  to  harvest  and 
pouch  several  seeds  from  the  same  pine  cone  and  visit 
more  than  one  cone  from  the  same  tree  or  genet  (if  a  clus- 
ter) (Tomback  1988;  Tomback  and  Knowles  1989),  they 
tend  to  place  sibling  seeds  in  the  same  seed  cache.  Dur- 
ing a  bumper  whitebark  pine  cone  crop  (1989)  in  Yellow- 
stone National  Park,  Tomback  and  Knowles  (1989)  ob- 
served nutcrackers  taking  entire  pouchloads  of  28  to  97 
seeds  {x  =  53  seeds,  n  =  8)  from  a  single  cone  or  individual 
tree  (genet)  or  trimk. 

One  consequence  of  this  behavior  is  the  tendency  for 
individuals  in  tree  clusters  to  be  genetic  relatives — either 
half-siblings,  siblings,  or  the  products  of  selflng.  Furnier 
and  others  (1987)  found  that  individuals  within  tree  clus- 
ters of  two  populations  were  genetically  more  similar 
(smaller  genetic  distance)  than  individuals  in  different 
tree  clusters.  Individuals  within  clusters  differed  by  a 
mean  of  1.66  alleles;  individuals  of  different  clusters  dif- 
fered by  a  mean  of  5.82  alleles  or  greater  (of  11  gene  loci). 
Two  studies  examined  genetic  relatedness  within  limber 
pine  clusters  using  the  Quellar  and  Goodnight  (1989)  esti- 
mator (table  4):  For  tree  clusters  in  the  Pawnee  National 
Grasslands  population,  Schuster  and  Mitton  (1991)  deter- 
mined that  individuals  were  related  on  average  as  half- 
siblings.  In  addition,  they  determined  that  cluster  mem- 
bers that  had  fused  together  at  the  base  were  more  closely 
related  on  average  (r  =  0.35)  than  cluster  members  that 
had  not  fused  (r  =  0.05).  Carsey  and  Tomback  (1992)  cal- 
culated that  cluster  members  in  their  upper  subalpine 
population  (3,310  m  elevation)  were  related  on  average 
as  between  half  and  full  siblings;  members  of  different 
clusters  were  unrelated  on  average  (table  4). 


46 


Possible  Biological  Implications  of  Tree 
Clustering — Despite  the  fact  that  tree  clusters  occiir  com- 
monly in  populations  of  bird-dispersed  pines  and  include 
a  large  proportion  of  the  genets  in  the  population,  little  is 
known  of  the  biology  of  this  growth  form  (Tomback  and 
others  1993).  Several  papers  have  speculated  about  the 
potential  advantages  and  disadvantages  to  individuals 
growing  in  clusters  (Schuster  and  Mitton  1991;  Tomback 
and  Linhart  1990;  Tomback  and  others  1993). 

Individuals  in  clusters  may  have  higher  survivorship 
than  single  genets  vmder  certain  environmental  condi- 
tions because  of  morphological  or  physiological  benefits 
(Bullock  1981;  Keeley  1988;  Mattes  1982).  A  sturdy,  well- 
anchored  structure  might  better  resist  the  strong  winds 
and  mechanical  damage  of  higher  elevation  environments. 
Root  grafting  may  also  occur  among  cluster  members,  par- 
ticularly relatives  (Schuster  and  Mitton  1991),  resulting 
in  efficient  water  and  nutrient  acquisition  (Holtmeier 
1986).  In  addition,  since  nutcrackers  may  establish  pines 
at  some  distance  from  conspecifics,  for  example,  in  burned 
areas  or  disturbed  patches,  individuals  in  tree  clusters 
may  cross-pollinate  and  thus  avoid  selfing  (Tomback  and 
Linhart  1990).  This  might  produce  a  high  proportion  of 
sound  seeds,  particularly  if  clusters  consisted  of  both  rela- 
tives and  nonrelatives.  Finally,  both  Schuster  and  Mitton 
(1991)  and  Tomback  and  Linhart  (1990)  point  out  that  the 
occurrence  of  kin  groups  of  bird-dispersed  pines  may  re- 
sult in  kin  selection  as  an  evolutionary  force.  Possible 
consequences  for  cluster  members  may  include  some  de- 
crease in  competition  or  facilitation  of  rootgrafting. 

With  respect  to  the  potential  disadvantages  to  indivi- 
duals growing  in  clusters,  one  area  of  possible  fitness  loss 
is  reproductive  output.  For  limber  pine  trees  in  three 
stands  in  the  Colorado  Front  Range  (same  populations 
studied  by  Carsey  and  Tomback  1992),  cluster  members 
produced  fewer  male  and  female  cones  than  did  single- 
tnmk  trees  during  each  of  3  years  (Feldman  1991; 
Feldman  and  Tomback  1991).  Reduction  of  reproductive 
output  may  result  from  competition  among  cluster  mem- 
bers for  moisture,  nutrients,  and  space.  These  effects  may 
be  exacerbated  for  limber  pine,  which  typically  grows  on 
windy,  xeric  sites  with  poor  soils  in  the  Colorado  Front 
Range  (Feldman  1991;  Feldman  and  Tomback  1991). 
Feldman  (1991)  and  Feldman  and  Tomback  (1991)  noted 
that  trees  in  clusters  had  smaller  diameters  than  did 
similarly  aged  single-trunk  trees,  supporting  the  possibil- 
ity of  competition. 

Another  possible  disadvantage  to  cluster  members  is 
that  their  seeds  may  be  more  inbred  than  seeds  produced 
by  single-genet  trees,  because  of  pollen  exchange  within 
clusters  (Furnier  and  others  1987;  Tomback  and  Linhart 
1990).  PoUtov  and  Krutovskii  (1990)  provide  some  evi- 
dence for  this  in  Siberian  stone  pine  (Pinus  sibirica), 
which  is  dispersed  by  the  Eurasian  nutcracker.  Examin- 
ing the  mating  system  of  Siberian  stone  pine  in  seven 
populations,  they  noted  an  excess  of  homozygotes  among 
seed  embryos  in  the  majority  of  their  samples.  They  at- 
tributed this  to  a  relatively  high  rate  of  self-pollination, 
but  pollination  among  related  cluster  members  would 
have  a  similar  effect. 


Genetic  Spatial  Structure  Within  Popvilations — 

Several  studies  indicate  that  in  many  wind-dispersed 
conifers  there  is  family  structure  within  popiilations 
(local  aggregations  of  genetic  relatives)  (see,  for  example, 
Knowles  1984;  Linhart  1989;  Linhart  and  others  1981). 
This  is  the  consequence  of  seed  dispersal  typically  re- 
stricted to  short  distances  (Ftunier  and  others  1987; 
Tomback  and  Linhart  1990),  and  the  availability  of  local, 
disturbed  patches  for  colonization  (see,  for  example, 
Linhart  and  others  1981).  In  contrast,  Furnier  and  others 
(1987)  did  not  find  any  evidence  of  family  structure 
among  the  neighboring  tree  clusters  in  two  whitebark 
pine  populations  (although  cluster  members  tended  to 
be  related).  They  attributed  this  lack  of  family  structiire 
with  respect  to  geographic  distance  to  the  seed-caching 
behavior  of  Clark's  nutcrackers.  Since  Clark's  nutcrack- 
ers do  not  defend  caching  territories,  several  different 
nutcrackers  may  cache  in  one  area,  often  bringing  seeds 
in  from  different  stands  of  trees  (Tomback  1978,  1982). 
This  probably  results  in  a  haphazard  distribution  of  re- 
lated seed  caches  (Furnier  and  others  1987;  Tomback 
and  Linhart  1990).  The  Eurasian  nutcracker  in  Sweden 
(Swanberg  1956)  and  in  the  Alps  (Mattes  1982)  stores 
many  of  its  caches  within  territories  defended  by  pairs. 
However,  the  seeds  cached  in  these  territories  are  both 
from  local  and  more  distant  parent  trees  (up  to  15  km. 
Mattes  1982),  so  there  is  also  considerable  mixing  in  any 
area.  PoHtov  and  Krutovskii  (1990)  examined  seven  popu- 
lations of  Siberian  stone  pine  and  also  foimd  no  evidence 
for  intrapopulation  substructure. 

GENETIC  DIVERSITY 

Generally  speaking,  conifers  show  less  genetic  diversity 
among  popiilations  (mean      =  0.068)  than  do  other  kinds 
of  plants  (mean  Gg^  =  0.250)  (Hamrick  and  Godt  1990). 
After  an  extensive  review  of  the  plant  literature,  Hamrick 
and  Godt  (1990)  concluded  that  animal-dispersed  species 
have  more  differentiated  populations  (higher  Gg^  values) 
than  do  wind-dispersed  species.  But,  seed  dispersal  by 
nutcrackers,  which  occurs  routinely  over  large  distances, 
might  actually  result  in  higher  levels  of  gene  flow  than 
from  seed  dispersal  by  wind  and  thus  lower  levels  of  popu- 
lation differentiation. 

Few  studies  have  examined  interpopulation  variation  in 
bird-dispersed  pines.  Schuster  and  others  (1989)  geneti- 
cally analyzed  two  limber  pine  populations  at  extreme 
ends  of  a  1,700-m  elevational  transect  in  the  Colorado 
Front  Range.  They  determined  that  pollen  flow  along  this 
gradient  was  impeded  by  pollination  phenology;  popula- 
tions separated  by  more  than  400  m  elevation  did  not 
have  overlapping  pollination  periods.  For  the  populations 
at  each  end  of  the  transect,  allele  frequencies  differed  sig- 
nificantly at  eight  of  the  10  loci  studied.  However,  the 
calculated  nimiber  of  migrants  exchanged  between  the 
populations  per  generation  (A^^  =  11.1)  was  well  above  the 
level  required  to  overcome  genetic  drift.  Given  the  very 
different  environmental  conditions  of  the  two  populations, 
Schuster  and  others  (1989)  concluded  that  natural  selec- 
tion rather  than  lack  of  gene  flow  may  have  differentiated 
the  populations.  They  also  suggested  that  seed  dispersal 


47 


by  Clark's  nutcrackers  is  likely  to  be  "a  more  important 
component  of  gene  flow  in  limber  pine  than  pollen  trans- 
fer, at  least  across  elevational  gradients." 

Schuster  and  Mitton  (1992)  also  compared  five  popula- 
tions of  limber  pine  across  a  100-km  range  and  found  low 
levels  of  differentiation  (mean  Gg^  =  0.035).  Hamrick  (un- 
published data  cited  in  Betancourt  and  others  1991)  ex- 
amined eight  populations  of  pinyon  pine  (Pinus  edulis) 
and  found  a  slightly  greater  differentiation  (mean  Gg^  = 
0.077). 

Nine  populations  of  the  Siberian  stone  pine  were  stud- 
ied by  Krutovskii  £ind  others  (1989)  to  examine  interpopu- 
lation  genetic  differentiation.  Their  results  indicated  only 
slight  differences  among  populations.  Using  Nei's  meas- 
ures of  diversity,  they  determined  that  on  average  about 
98  percent  of  allozyme  variation  was  attributed  to  intra- 
population  variability  and  only  2  percent  to  interpopula- 
tion  variability  (mean  Gg^  =  0.022).  However,  they  did 
find  a  significant  correlation  between  geographical  dis- 
tance and  genetic  differences  among  populations,  but  only 
for  populations  separated  by  400  km  or  more.  They  also 
found  a  higher  level  of  differentiation  among  populations 
within  the  same  region  than  between  regions,  which  may 
reflect  adaptation  to  local  environmental  conditions.  The 
high  level  of  intrapopulational  variation  in  Siberian  stone 
pine  is  very  similar  to  what  has  been  found  for  most  wind- 
dispersed  conifer  species  (Krutovskii  and  others  1989). 

EVOLUTIONARY  IMPLICATIONS 

Tomback  and  Linhart  (1990)  speculated  that  bird- 
dispersed  pines  might  be  more  prone  to  genetic  drift  than 
are  wind-dispersed  species.  The  ranges  of  these  pines, 
particularly  subalpine  species,  tend  to  be  highly  frag- 
mented, and  small,  isolated  populations  are  not  uncom- 
mon. These  smaller  populations  may  show  founder  ef- 
fects, particularly  loss  of  rare  alleles.  In  addition,  because 
of  the  occurrence  of  tree  clusters  and  the  potential  for 
cross-pollination  among  cluster  members,  there  may  be 
high  rates  of  inbreeding  in  these  populations.  When  effec- 
tive population  size  is  smaller  than  100,  genetic  drift  can 
be  far  more  potent  in  causing  gene  frequency  changes 
than  natural  selection  (Futuyma  1986  and  references 
therein).  It  is  possible  that  the  evolution  of  cone,  seed, 
and  tree  morphology  in  bird-dispersed  pines  was  in  part 
the  consequence  of  new  traits  that  arose  from  a  combina- 
tion of  drift  and  inbreeding  in  small,  isolated  populations. 
Nutcrackers,  or  ancestral  corvids,  selected  for  preferred 
traits  and  spread  these  traits  to  other  populations  or 
founded  new  populations  with  these  traits. 

If  bird-dispersed  pines  have  this  population  structure, 
Tomback  and  Linhart  (1990)  make  two  predictions:  "(1) 
there  is  greater  genetic  variation  among  populations  of 
bird-dispersed  pines  than  among  wind-dispersed  pines, 
and  (2)  there  are  small,  isolated  populations  of  bird- 
dispersed  pines  with  highly  differentiated  gene  pools  com- 
pared to  one  another  and,  particularly,  compared  to  large 
populations  that  may  regularly  experience  some  gene  flow 
from  other  large  populations." 

Although  the  earliest  populations  of  bird-dispersed  pines 
may  have  had  the  characteristics  described  by  Tomback 


and  Linhart  (1990),  the  findings  of  Krutovskii  and  others 
(1989),  Schuster  and  Mitton  (1992),  and  Hamrick  (unpub- 
lished data  cited  in  Betancourt  and  others  1991)  do  not 
support  the  first  prediction,  and  it  may  well  be  that  the 
great  distances  over  which  nutcrackers  disseminate  seeds 
provide  effective  gene  flow.  Betancourt  and  others  (1991) 
describe  a  disjunct  population  of  Colorado  pinyon  pine  in 
northern  Colorado  that  was  the  result  of  a  founding  event 
probably  about  400  years  ago.  Compared  to  other  pinyon 
pine  populations,  the  average  percentage  of  polymorphic 
loci  and  nimiber  of  alleles  per  locus  were  reduced,  although 
levels  of  heterozygosity  were  not  affected.  These,  how- 
ever, are  typical  founder  effects  (Betancourt  and  others 
1991;  Futuyma  1986).  Within  the  last  century,  the  found- 
ing population  has  given  rise  to  four  descendant  popula- 
tions, three  with  fewer  than  50  trees  each  (Premoli  and 
others  1993).  The  two  smallest,  most  distant  descendant 
populations  had  fewer  alleles  per  polymorphic  locus  and 
reductions  in  individual  heterozygosity  compared  to  the 
larger  populations.  The  mean  Gg^  value  of  0.060  indicated 
little  differentiation  among  populations.  Because  tree 
clusters  do  not  occur  often  in  Colorado  pinyon  pine,  these 
new  populations  may  not  adequately  test  prediction  2. 
However,  they  do  demonstrate  that  seed  dispersal  by 
birds  results  in  small,  isolated  foimding  populations. 
Much  more  work  is  required  to  thoroughly  characterize 
the  genetic  population  structure  of  bird-dispersed  pines 
and  to  address  questions  concerning  the  biology  of  tree 
clusters. 

ACKNOWLEDGMENTS 

We  thank  Konstantin  Krutovskii  and  Dmitri  Politov  for 
providing  translations  of  their  papers  £ind  for  interesting 
discussions  during  the  workshop.  In  addition,  we  are 
grateful  to  Jeff  Karron,  Andrea  Premoli,  and  Kathleen 
Shea  for  helpful  comments  on  an  earlier  draft  of  the  pa- 
per. Support  for  DFT  for  attending  the  SSPE  Workshop 
was  provided  by  the  College  of  Liberal  Arts  and  Sciences, 
University  of  Colorado  at  Denver. 

REFERENCES 

Betancourt,  J.  L.;  Schuster,  W.  S.;  Mitton,  J.  B.; 
Anderson,  R.  S.  1991.  Fossil  and  genetic  history  of 
a  pinyon  pine  (Pinus  edulis)  isolate.  Ecology.  72(5): 
1685-1697. 

Bibikov,  D.  I.  1948.  On  the  ecology  of  the  nutcracker. 
Trudy  Pechorskogo-Ilychskogo  Gosudarstvennogo 
Zapovednika.  4(2):  89-112.  [In  Russian,  translated 
by  L.  Kelso]. 

Bullock,  S.  J.  1981.  Aggregations  of  Prunus  ilicifolia 
(Rosaceae)  during  dispersal  and  its  effect  on  survival 
and  growth.  Madrono.  28:  94-95. 

Carsey,  K.  S.;  Tomback,  D.  F.  1992.  Growth  form  distribu- 
tion and  genetic  relationships  in  tree  clusters  of  Pinus 
flexilis,  a  bird-dispersed  pine.  Manuscript. 

Critchfield,  W.  F.;  Little,  E.  L.,  Jr.  1966.  Geographic 
distribution  of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture,  For- 
est Service.  97  p. 


48 


Epperson,  B.  K  1990.  Spatial  patterns  of  genetic  varia- 
tion within  plant  populations.  In:  Brown,  A.  H.  D.; 
Clegg,  M.  T.;  Kahler,  A.  L.;  Weir,  B.  S.,  eds.  Plant  popu- 
lation genetics,  breeding,  and  genetic  resources. 
Sunderland,  MA:  Sinauer  Associates:  229-253. 

Feldman,  R.  1991.  Growth  form  and  reproductive  output 
in  limber  pine:  the  cost  of  mutualism.  Denver,  CO:  Uni- 
versity of  Colorado  at  Denver,  Department  of  Biology. 
46  p.  Thesis. 

Feldman,  R.;  Tomback,  D.  F.  1991.  Growth  form  and  re- 
productive output  in  limber  pine:  the  cost  of  mutualism. 
Manuscript. 

Fumier,  G.  R.;  Knowles,  P.;  Clyde,  M.  A.;  Dancik,  B.  P. 
1987.  Effects  of  avian  seed  dispersal  on  the  genetic 
population  structure  of  whitebark  pine  populations. 
Evolution.  4(3):  607-612. 

Futuyma,  D.  J.  1986.  Evolutionary  biology.  2d  ed. 
Sunderland,  MA:  Sinauer  Associates.  600  p. 

Hamrick,  J.  L.;  Godt,  M.  J.  W.  1990.  Allozyme  diversity  in 
plant  species.  In:  Brown,  A.  H.  D.;  Clegg,  M.  T.;  Kahler, 
A.  L.;  Weir,  B.  S.,  eds.  Plant  population  genetics,  breed- 
ing, and  genetic  resoiirces.  Simderland,  MA:  Sinauer 
Associates:  43-63. 

Hayashida,  M.  1989.  Seed  dispersal  by  red  squirrels  and 
subsequent  establishment  of  Korean  pine.  Forest  Ecol- 
ogy and  Management.  28:  115-129. 

Holtmeier,  F.  -K.  1986.  Uber  Bamninseln  (Kollektive) 
an  der  klimatischen  Waldgrenze — imter  besonderer 
Beriicksichtigung  von  Beobachtungen  inverschiedenen 
Hochgebirgen  Nord-amerikas.  Wetter  und  Leben.  38: 
121-139. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Keeley,  J.  E.  1988.  Population  variation  in  root  grafting 
and  a  hypothesis.  Oikos.  52(3):  364-366. 

Kephart,  S.  R.  1990.  Starch  gel  electrophoresis  of  plant 
isozymes:  a  comparative  analysis  of  techniques.  Ameri- 
can Journal  of  Botany.  77(5):  693-712. 

Knowles,  P.  1984.  Genetic  v£iriability  among  and  within 
closely  spaced  populations  of  lodgepole  pine.  Canadian 
Journal  of  Genetics  and  Cytology.  26:  177-184. 

Krutovskii,  K.  V.;  Politov,  D.  V.;  Altukhov,  Y.  P.;  Milyutin, 
L.  I.;  Kuznetsova,  G.  V.;  Iroshnikov,  A.  I.;  Vorob'ev,  V.  N.; 
Vorob'eva,  N.  A.  1989.  Genetic  variability  in  the  Sibe- 
rian stone  pine,  Pinus  sibirica  Du  Tour.  IV.  Genetic 
diversity  and  the  extent  of  genetic  differentiation  be- 
tween populations.  Genetika.  25(11):  2009-2032.  [Trans- 
lation published  by  Pleniun  Publishing  Corp.  in  1991]. 

Laimer,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in 
the  ecology  and  evolution  of  limber  and  whitebark 
pines.  In:  Dancik,  B.  P.;  Higginbotham,  K.  0.,  eds.  Sixth 
North  Americsin  forest  biology  workshop:  proceedings. 
Edmonton,  AB:  University  of  Alberta:  15-48. 

Lepper,  M.  G.  1974.  Pinus  flexilis  James  and  its  environ- 
mental relationships.  Davis,  CA:  University  of  California 
at  Davis,  Graduate  Group  in  Ecology.  197  p.  Dissertation. 

Ligon,  J.  D.  1978.  Reproductive  interdependence  of 
pinon  jays  and  pinon  pines.  Ecological  Monographs. 
48:  111-126. 


Linhart,  Y.  B.  1989.  Interactions  between  genetic  and  eco- 
logical patchiness  in  forest  trees  and  their  dependent 
species.  In:  Bock,  J.  E.;  Linhart,  Y.  B.,  eds.  Evolutionary 
ecology  of  plants.  Boulder,  CO:  Westview  Press:  1-31. 

Linhart,  Y.  B.;  Mitton,  J.  B.;  Sturgeon,  K.  B.;  Davis,  M.  L. 
1981.  Genetic  variation  in  space  and  time  in  a  popula- 
tion of  ponderosa  pine.  Heredity.  46:  407-426. 

Linhart,  Y.  B.;  Tomback,  D.  F.  1985.  Seed  dispersal  by 
nutcrackers  causes  multi-trunk  growth  form  in  pines. 
Oecologia.  67:  107-110. 

Mattes,  H.  1982.  Die  Lebengemeinschaft  von  Tannenhaher 
imd  Arve.  Berichte  Eidenossische  Anstalt  fur  das 
forstliche  Versuchswesen,  Birmensdorf.  Nr.  241.  74  p. 

McCaughey,  W.  W.;  Schmidt,  W.  C;  Shearer,  R.  C.  1986. 
Seed-dispersal  characteristics  of  conifers  in  the  Inland 
Mountain  West.  In:  Shearer,  R.  C,  comp.  Proceedings — 
conifer  tree  seed  in  the  Inland  Mountain  West  sympo- 
sium; 1985  August  5-6;  Missoula,  MT.  Gen.  Tech.  Rep. 
INT-203.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Research  Station:  50-62. 

McNaughton,  G.  M.  1984.  Comparative  water  relations 
of  Pinus  flexilis  at  high  and  low  elevations  in  the  cen- 
tral Rocky  Mountains.  Laramie,  WY:  University  of 
Wyoming,  Department  of  Botany.  59  p.  Thesis. 

Mezhenny,  A.  A.  1961.  Food  competitors,  enemies  and  dis- 
eases. In:  Egorov,  O.  V.,  ed.  Ecology  and  economics  of 
the  Yakut  squirrel.  Moscow:  Akademiya  Nauk:  124-129. 

Politov,  D.  v.;  Krutovskii,  K.  V.  1990.  Genetic  variation  of 
Siberian  stone  pine  Pinus  sibirica  Du  Tom*.  V.  Analysis 
of  mating  system.  Genetika.  26(11):  2002-2011.  [Trans- 
lation published  by  Plenimi  Publishing  Corp.  in  1991]. 

Premoh,  A.  C;  Chischilly,  S.;  Mitton,  J.  B.  1993.  Levels 
of  genetic  variation  captured  by  four  descendant 
populations  of  pinyon  pine  (Pinus  edulis  Engelm.). 
Manuscript. 

Quellar,  D.  C;  Goodnight,  K.  F.  1989.  Estimating  relate d- 
ness  using  genetic  markers.  Evolution.  43:  258-275. 

Schuster,  W.  S.  F.;  Alles,  D.  L.;  Mitton,  J.  B.  1989.  Gene 
flow  in  limber  pine:  evidence  from  pollination  phenology 
and  genetic  differentiation  along  an  elevational  transect. 
American  Journal  of  Botany.  76(9):  1395-1403. 

Schuster,  W.  S.  F.;  Mitton,  J.  B.  1991.  Relatedness  within 
clusters  of  a  bird-dispersed  pine  and  the  potential  for 
kin  interactions.  Heredity.  67:  41-48. 

Schuster,  W.  S.  F.;  Mitton,  J.  B.  1992.  Gene  flow  in 
limber  pine.  II.  Data  for  five  populations.  Manuscript 
in  preparation. 

Swanberg,  P.  O.  1956.  Territory  in  the  thick-billed  nut- 
cracker A^uci/ra^a  caryocatactes.  Ibis.  98:  412-419. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. The  Living  Bird.  16:  123-161. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.  1988.  Nutcracker-pine  mutualisms:  multi- 
tnmk  trees  and  seed  size.  In:  Ouellet,  H.,  ed.  Acta  XIX 
Congressus  Internationalis  Ornithologici,  Vol.  1;  1986 
June  22-29.  Ottawa,  ON:  University  of  Ottawa  Press: 
518-527. 


49 


Tomback,  D.  F.  1988.  [Unpublished  observations  in  the 
Upper  Engadine  Valley  of  Switzerland,  October  1988]. 

Tomback,  D.  F.;  Knowles,  J.  W.  1989.  Post-fire  whitebark 
pine  seed  dispersal  by  Clark's  nutcracker  in  Yellowstone 
National  Park.  Unpublished  data  and  observations. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Coevo- 
lution  of  whitebark  pine  and  nutcrackers:  implications 
for  forest  regeneration.  In:  Schmidt,  W.  C;  McDonald, 
K  J.,  comps.  Proceedings — sjnnposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  118-129. 

Tomback,  D.  F.;  Holtmeier,  F.  -K.;  Mattes,  H.;  Carsey, 
K.  S.;  Powell,  M.  L.  1993.  Tree  clusters  and  growth  form 


distribution  in  Pinus  cembra,  a  bird-dispersed  pine. 

Arctic  and  Alpine  Research.  [In  press]. 
Turcek,  F.  J.  1966.  Uber  das  Wiederauffinden  von  im 

Boden  versteckten  Samen  durch  Tannen-und 

Eichelhaher.  Waldhygiene.  6:  215-217. 
Vander  Wall,  S.  B.  1992.  The  role  of  animals  in  dispersing 

a  "wind-dispersed"  pine.  Ecology.  73:  614-621. 
Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations 

of  Clark's  nutcracker  and  the  pinon  pine  for  efficient 

seed  harvest  and  dispersal.  Ecological  Monographs. 

47:  89-111. 

Weaver,  T.;  Jacobs,  J.  1990.  Occvirence  of  multiple  stems  in 
whitebark  pine.  In:  Schmidt,  W.  C;  McDonald,  K  J., 
comps.  Proceedings — symposium  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
156-159. 


50 


Basic  Ecology 


International  Workshop 
St.  Moritz  1 992 


EXPERIENCES  WITH  REPRODUCTION 
OF  CEMBRA  PINE 


Ernst  Frehner 
Walter  Schonenberger 


Abstract — ^As  regards  reproduction,  cembra  pine  {Pinus  cembra 
L.)  is  a  special  case.  If  the  greater  part  of  the  seeds  are  stolen  by 
nutcrackers,  they  should  be  harvested  before  they  are  ripe  and 
then  matured.  With  correct  treatment  it  is  possible  to  store 
seeds  for  up  to  7  years.  After  soaking  in  water,  germination  be- 
gan almost  at  once.  The  separate  steps  along  the  long  process 
from  harvesting  via  seed  extraction,  maturation,  storage,  testing, 
sowing,  and  transplantation  to  planting  out  are  described  in 
detail. 


in  the  Alps,  and  that  the  existing  stands  would  diminish 
considerably  within  a  few  generations. 

Depending  on  the  number  of  cones  produced,  the  nut- 
cracker stores  seeds  as  winter  provisions  in  nimierous 
caches,  usually  numbering  over  10,000,  each  containing 
from  three  to  five  seeds,  but  only  consimies  about  80  per- 
cent of  its  reserve.  The  remaining  seeds  contribute 
greatly  to  the  regeneration  of  cembra  pine  stands.  The 
nutcracker  may  transport  seeds  over  a  distance  of  15  km 
or  more  and  over  a  difference  in  height  of  up  to  600  m. 


This  study  reports  our  experiences  over  many  years  of 
seed  £ind  afforestation  trials  with  cembra  pine  {Pinus 
cembra  L.),  from  seed  harvesting  to  planting. 

Cembra  pine  is  a  special  case  as  regards  reproduction 
(Frehner  and  Fuerst  1992).  The  extreme  conditions  in  its 
natural  environment  at  the  upper  timberline  strongly  in- 
fluence flowering,  fertilization,  and  seed  maturation,  so 
that  the  quantity  and  quality  of  seed  varies  from  year  to 
year  (table  1).  The  results  of  repeated  studies  on  cone 
quality,  maturation,  and  storage  vary  correspondingly. 

The  special  features  of  reproduction  in  cembra  pine  be- 
gin with  seed  collection.  Except  in  full  mast  years,  most 
of  the  seeds  have  been  eaten  or  collected  by  the  nut- 
cracker by  the  time  of  their  natural  maturation.  Like  the 
seeds  of  many  other  tree  and  shrub  species,  those  of 
cembra  pine  require  several  months  of  dormancy.  Matu- 
ration of  the  seed  and  time  of  extraction  is  therefore  very 
important  for  the  development  of  the  embryo.  To  better 
exploit  fiill  mast  years,  the  durability  of  the  seeds  should 
be  extended  through  improved  storage  methods.  It  is  also 
important  that  stored  seeds  should  be  induced  to  germi- 
nate and  produce  successful  saplings  at  short  notice.  Also 
in  terms  of  sowing,  transplantation,  and  potting,  cembra 
pine  requires  special  conditions. 

SYMBIOSIS  WITH  THE 
NUTCRACKER 

The  importance  of  the  nutcracker  for  the  natural  distri- 
bution of  cembra  pine  is  undisputed.  The  detailed  study 
by  Mattes  (1982)  on  this  relationship  concludes  that  with- 
out the  nutcracker  there  would  be  no  cembra  pine  forests 


CONE  COLLECTION 

In  years  with  normal,  moderate  fructification,  the  nut- 
cracker begins  to  pick  the  seeds  at  medium  altitudes  as 
soon  as  ripening  has  begun,  about  mid-August,  and  con- 
tinues to  do  so  until  about  the  end  of  September.  This 
means  that  collecting  cones  is  a  race  with  the  bird.  If 
cones  are  to  be  collected  in  years  with  partial  mast,  this 
must  be  done  at  an  early  stage,  while  they  are  still  un- 
ripe, before  the  birds  begin  their  robbery,  and  then  stored 
under  optimvmi  conditions  for  maturation. 

In  long-term  observations,  19  percent  of  the  years  pro- 
duced full  mast,  55  percent  partial  mast,  and  26  percent 
failed  to  produce  (Mattes  1982).  As  a  rule  it  is  only  in  fiill 
mast  years  that  enough  seeds  ripen  on  the  tree  to  meike 
harvesting  worthwhile;  in  this  case  the  nutcracker  does 
not  have  time  to  devour  or  carry  away  all  the  remaining 
seeds. 

Neither  color  nor  size  of  cones  has  any  influence  on  seed 
quality  or  planting  success.  The  costs  of  harvesting  and 
seed  extraction  are  considerably  greater  for  small  cones. 

With  early  harvesting  it  is  very  important  that  the 
seeds  should  be  allowed  to  ripen  imder  optimum  condi- 
tions of  moisture  and  warmth,  as  the  embryos  are  in  a 
very  early  stage  of  development  (only  up  to  3  mm  long). 

Studies  on  protected  cones  attached  to  mother  trees  at 
2,000  m  have  shown  that  the  length  of  the  embryo  hardly 
changes  between  mid-September  and  the  end  of  October. 
In  contrast,  embryos  in  cones  harvested  early  (mid- 
September)  have  elongated  by  3-4  mm.  This  shows  that 


Table  1 — Characteristics  of  cembra  pine  seed  crops 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11, 1992. 

Ernst  Frehner  is  former  Head  of  the  Research  Nursery,  Walter 
Schonenberger  is  Head  of  the  Research  Section  Silviculture,  Swiss 
Federal  Institute  for  Forest,  Snow  and  Landscape  Research,  CH-8903 
Birmensdorf,  Switzerland. 


Harvesting  time  per  tree 

Yield  per  tree 

Number  of  cones  per  tree 

Weight  per  cone 

Seed  yield  (fresh  weight) 

Number  of  living  seeds  per  cone 

Thousand-seed  weight 


about  1  hour 
about  15  kg  cones 
about  150-200 
60-100  g 
20-30  percent 
50-70 
300-400  g 


52 


Table  2 — Cone  collection,  maturation,  and  seed  extraction 

September  Cone  harvesting  (embryo  length  up  to  3  mm) 

September  Stratification  of  cones  for  maturation  at  1 0-20  °C 

for  8-1 0  weeks 

November  Seed  extraction  (embryo  length  3-5  mm) 


it  is  not  worthwhile  enclosing  whole  trees  in  nets  or  treat- 
ing the  cones  with  chemicals.  Early  harvesting  and  favor- 
able conditions  of  stratification  are  preferable  to  late 
harvesting. 

CONE  MATURATION 

Stratification  (storing  cones  in  layers  in  a  moist,  airy 
substrate)  is  the  simplest  and  most  certain  technique  of 
accelerating  ripening  and  reducing  inhibition  of  germi- 
nation so  that  the  seeds  are  ready  to  germinate  quickly. 
Stratification  simulates  the  natural  conditions  of  moisture 
and  warmth  prevailing  in  the  forest,  or  may  even  be 
better. 

The  fi-eshly  harvested  cones  are  stored  in  layers  at  a 
proportion  of  one  part  cones  to  five  parts  substrate.  Suit- 
able substrates  are  washed  sand  or  a  mixture  of  sand  and 
peat.  The  substrate  must  be  sieved  so  that  its  particles 
are  smaller  than  the  seeds.  The  stratified  cones  should 
then  be  stored  in  a  cupboardlike  container  with  good 
drainage  that  protects  the  cones  from  attack  by  mice  or 
birds. 

The  development  of  the  embryos  depends  greatly  on  the 
temperature  of  the  substrate  and  the  water  content  of  the 
cones.  Studies  have  shown  that  the  embryos  develop  only 
slowly  at  low  temperatures,  while  high  temperatures  also 
have  a  negative  influence.  The  best  results  were  obtained 
with  substrate  temperatures  alternating  between  10  and 
20  °C.  It  is  advisable  to  keep  the  stratified  cones  in  a  pro- 
tected place  (for  example,  an  unheated  plastic  tiinnel  or 
glasshouse),  where  the  temperature  may  rise  toward 
25  °C  during  the  day  and  sink  to  about  5  °C  at  night. 
Naturally,  the  temperatiire  within  the  substrate  does  not 
vary  so  greatly. 

Moisture  is  very  important.  The  substrate  should  be 
moist  and  airy,  and  the  cones  should  be  just  moist  enough 
to  prevent  the  seed  coat  from  shrinking  away  from  the 
endosperm.  Maturation  to  the  point  of  seed  extraction 
proceeds  much  better  under  moist  conditions  than  with 
dry  storage.  With  ripening  under  dry  conditions  the  ger- 
mination rate  is  50  percent  lower  than  that  after  moist 
storage,  while  dormancy  is  10  times  longer. 

SEED  EXTRACTION 

After  8  to  10  weeks'  stratification,  the  seeds  can  be  ex- 
tracted. The  cones,  which  are  by  this  time  beginning  to 
disintegrate,  are  sieved  out  of  the  substrate,  left  to  dry  for 
a  while,  and  then  rubbed  through  finer  sieves  to  separate 
the  seeds  from  scales  and  cone  axes.  The  finer  scales  and 
substrate  particles  can  then  be  removed  through  further 


sieving  and  washing  in  a  water  bath,  either  immediately 
or  after  another  ripening  process.  In  a  water  bath  the  im- 
fertilized  seeds  (empty  seed  cases)  float  to  the  top  and  can 
be  decanted.  To  summarize,  the  course  of  cone  matura- 
tion is  shown  in  table  2. 

SEED  MATURATION 

The  treatment  of  the  seeds  after  extraction  depends  on 
their  intended  use  and  how  long  they  are  to  be  stored. 

Under  natural  conditions  it  takes  2  to  3  years  for  the 
seeds  to  germinate  after  the  cones  have  fallen.  This  may 
be  due  to  many  factors,  for  example,  hard  and  imperme- 
able seed  cases,  incomplete  development  of  the  embryo,  or 
inhibiting  substances  in  the  endosperm,  in  the  seed  case, 
or  elsewhere  in  the  seed.  In  cembran  pine,  the  main  rea- 
son is  the  low  stage  of  development  of  the  embryos:  at  the 
time  of  harvesting  the  embryos  are  not  more  than  3  mm 
long. 

With  artificial  maturation  using  variable  temperatures, 
the  period  between  harvesting  and  germination  can  be  re- 
duced to  one  winter.  Here,  the  aim  is  to  induce  the  em- 
bryos to  grow  fi"om  a  length  of  3  mm  or  less  to  the  germi- 
nation length  of  7-9  mm  (table  3).  To  promote  embryo 
growth  even  after  seed  extraction,  the  seeds  are  alter- 
nately exposed  to  warmth  and  cold  in  a  monthly  rotation. 
Exposure  to  warmth  for  periods  of  more  than  4  to  6  weeks 
does  not  positively  influence  embryo  development.  Great 
attention  should  be  paid  to  water  content  throughout  the 
whole  period  of  seed  treatment,  as  during  cone  storage. 
The  better  the  seed  maturation,  the  better  the  embryo  de- 
velopment and,  consequently,  the  better  the  seeds  are 
suited  for  drying  and  storage. 

BASIC  PRINCIPLES  OF  STORAGE 

After  drying,  the  seeds  are  placed  in  one  of  the  con- 
tainer types  mentioned  and  stored  imder  cool  conditions. 
In  the  early  stages  they  should  be  regularly  inspected. 

Water  content: 

•  the  lower  the  water  content,  the  lower  the  storage 
temperature; 

•  the  lower  the  storage  temperature,  the  longer  the 
seeds  can  be  stored; 

•  the  longer  the  seeds  have  been  stored,  the  longer 
they  must  be  stratified  before  sowing. 


Table  3 — Results  of  the  maturation  of  cembran  pine  seeds 
(harvested  September  22)  after  warm-cold-warm 
treatment 


Features 

September  23 

February  28 

Seed  length  (peeled) 

9.5  mm 

10.4  mm 

Channel  length 

4.5  mm 

8.5  mm 

Embryo  length 

1.0  mm 

8.4  mm 

Seeds  with  visible  embryos 

11  percent 

72  percent 

Seeds  ready  to  germinate 

96  percent 

76  percent 

53 


Storage  containers: 

•  at  high  water  content  (>20  percent),  in  perforated 
plastic  bags; 

•  at  low  water  content  (<15  percent),  in  airtight  con- 
tainers accommodating  2-5  kg. 

Storage  rooms:  climatized: 

•  for  short-term  storage  with  high  water  content, 
1-3  °C; 

•  for  fairly  long  storage  with  low  water  content, 
-5  to  -10  °C. 

Freeze  drying  has  proved  unsuitable  for  cembran  pine 
seeds,  because  the  water  content  is  too  high. 

Fungal  attack:  for  prevention,  the  seeds  should  be 
treated  with  a  fungicide  and  regularly  examined  as  to 
smell  and  condensed  moisture  indicating  a  too-high  water 
content  in  the  containers;  if  this  occurs,  the  seeds  should 
be  dried  immediately. 

STORAGE  OF  MATURE  SEEDS 

Without  treatment  for  reduction  of  water  content  the 
seeds  cannot  be  stored  longer  than  one  winter.  With  re- 
duction of  water  content  to  about  25  percent,  the  seeds 
can  be  stored  at  1-3  °C  for  up  to  3  years  without  great  loss 
in  the  germination  rate. 

Later  trials  showed  that  gently  handled  seeds  could  be 
dried  to  a  water  content  of  15  percent,  and  that  under 
storage  at  -5  to  -10  °C  the  germination  capacity  did  not 
decrease  for  up  to  7  years.  A  water  content  of  more  than 
20  percent  rendered  storage  below  0  °C  iinfeasible. 

THREE  SEED  TREATMENTS 

After  seed  extraction  three  treatments  for  three  differ- 
ent storage  periods  can  be  defined  as: 

•  Treatment  1:  for  immediate  sowing  without  storage 
(table  4). 

•  Treatment  2:  for  storage  over  2  to  3  years  (table  5). 

•  Treatment  3:  for  storage  over  3  to  7  years  (table  6). 

GERMINATION  TRIALS 

Germination  trials  begin  in  December.  The  seeds  are 
sown  in  a  sterile  peat/sand  in  sterilized  containers,  being 

Table  4 — Treatment  1 :  for  immediate  sowing  without  storage 

November  Seed  stratification  outdoors  for  5-6  months; 

substrate  three  parts  washed  sand  to  one  part 
seeds 

April  Sowing  (usually  30-50  percent  of  seeds  lie-over) 

Advantages:        No  need  for  climatization 

Disadvantages:    No  storage  possible,  annual  seed  harvesting 

necessary,  high  proportion  of  lie-over,  high  risk 
of  robbery  by  birds  and  mice 


Table  5 — Treatment  2:  for  storage  over  2-3  years  with  high  water 

content  under  refrigeration  in  plastic  bags  (for  example,  in 
years  with  partial  or  full  mast) 


November  Seed  stratification  for  4-6  weeks  at  1 6-1 8  °C  in 

washed  sand 

December/January  Sieving  and  cleaning  of  seeds;  drying  of  seeds 
for  1  -2  weeks  at  1 6-1 8  °C  to  about 
25  percent  water  content:  embryo  length 
increases  to  5-7  mm 

January  Storage  of  seeds  in  perforated  plastic  bags  at 

1  -3  °C  over  2-3  years 

October  Seeds  are  soaked  for  1  -2  days;  stratification 

outdoors  for  6  months  in  washed  sand 

April/May  Sowing  (no  lie-over  to  be  expected) 


Advantages:  Allows  storage  for  up  to  3  years  without  lie-over 

Disadvantages:       Storage  climatized  at  1-3  °C  necessary,  high 
water  content,  risk  of  rotting  and  fungal 
infection,  long  stratification  period 


normally  spread  and  pressed  down,  then  covered  with 
about  1  cm  sand  or  substrate.  The  trials  are  best  con- 
ducted in  rooms  without  direct  sunlight;  the  ideal  tem- 
perature is  16  to  18  °C.  Alternating  temperatures  have 
a  positive  effect.  The  seeds  should  be  watered  by  hand 
according  to  need.  Storage  in  greenhouses  with  automatic 
spraying  has  not  proved  suitable.  The  germination  tests 
take  6-8  weeks. 


Table  6 — Treatment  3:  for  storage  in  airtight  containers  for  3-7  years 
with  low  water  content  and  freezing  temperatures  (for 
example,  for  seeds  harvested  In  good  partial  or  full  mast 
years) 


December 

Seed  stratification;  first  warm  treatment  at 

16-18  °C  for  4-5  weeks 

January 

Cold  treatment  at  1  -3  °C  for  4-5  weeks 

February 

Second  warm  treatment  at  1 6-1 8  °C  for 

4-5  weeks  until  germination  begins;  embryo 

length  increases  to  7-9  mm 

February 

Sieving  and  cleaning  of  seeds;  drying  at 

1 6-1 8  °C  for  1  -2  weeks  tol  5  percent  water 

content 

March 

Storage  in  airtight  plastic  bags  or  glass 

containers  at  -5  to  -1 0  °C  for  3-7  years 

April 

Defrosting  for  1-2  days,  soaking  in  lukewarm 

water  for  1-2  days,  sowing 

Advantages: 

Seeds  can  be  stored  for  up  to  7  years,  no 

lie-over,  short-term  stratification  (6-8  weeks). 

good  mast  years  and  seed  quality  can  be 

fully  exploited 

Disadvantages: 

Long-term  deep-freezing  required 

54 


SOWING 

The  seeds  are  sown  between  the  end  of  April  and  mid- 
May.  The  number  of  seeds  to  be  sown  depends  on  the  re- 
sults of  the  germination  trials.  Because  germination  is 
slow,  the  seeds  should  be  sown  in  boxes  resistant  to  mice 
and  birds.  The  best  substrates  are  soils  without  lime- 
stone mixed,  according  to  conditions  and  pH,  with  peat 
(which  loosens  the  soil,  maintains  a  pH  between  5  and  6, 
and  regulates  the  moisture  level)  or  raw  humus  contain- 
ing mycorrhizal  fungi  from  cembran  pine  stands.  The 
seed  and  transplantation  beds  should  continuously  be 
occupied  by  cembran  pines  except  for  short  periods  with 
green  fertilizing. 

Cembran  pine  grows  best  in  thickly  sown  beds,  with 
1,200  to  1,800  seedlings  per  square  meter,  grown  for  2 
to  3  years  in  a  seedbed. 

If  the  seeds  begin  to  germinate  during  stratification, 
they  should  be  sown  immediately.  If  this  is  unfeasible, 
they  can  be  frozen  at  -3  to  -4  °C  and  sown  as  required. 
Seeds  that  have  already  germinated  and  whose  radicle  is 
longer  than  1  cm  should  not  be  planted  in  a  seedbed,  but 
thinned  out;  otherwise  the  resulting  plants  are  liable  to 
be  worthless,  with  tortuous  root  collars,  short  stems, 
small  crowns,  and  few  fine  roots.  Size,  quality,  and  yield 
are  definitely  better  in  thinned  seedlings.  Shortening  the 
roots  causes  deformations. 

We  did  not  run  any  trials  with  sowing  in  autumn,  as 
the  risks  of  attack  by  mice  and  birds  and  also  snow  pres- 
sure on  the  protection  lattice  are  then  very  high. 

TRANSPLANTING 

The  specially  designed  beds  are  continuously  supple- 
mented with  acid  soil  (for  example,  peat)  and  if  necessary 
with  chalk-free  acid  fertilizer.  To  hinder  invasion  by 
weeds  and  frost  heaving  and  to  restrict  evaporation,  and 


also  to  maintain  a  good  soil  structiire,  the  beds  are 
mulched. 

It  is  recommended  that  3-year-old  saplings  be  trans- 
planted in  spring  or  summer  (from  mid-July).  This  can  be 
done  by  machine,  with  a  spacing  of  10  by  18  cm.  Saplings 
with  poor  root  development  should  be  discarded,  as  it  is 
not  worth  planting  them  out.  The  saplings  stay  in  the 
transplanting  bed  for  3  years. 

POTTING 

High-altitude  afforestations  are  exposed  to  high  risks 
from  extreme  weather  conditions,  snow,  and  pests.  Expe- 
rience has  shown  that  seedlings  in  peat-fiber  pots  succeed 
better  than  those  with  naked  roots,  especially  on  diffic\ilt 
sites,  so  potting  is  recommended.  The  production  of  pot- 
ted plants  has  been  described  by  Schonenberger  and 
others  (1990).  It  is  important  to  keep  the  pots  well 
watered.  The  saplings  should  only  be  planted  out  when 
the  root  tips  begin  to  penetrate  through  the  peat-fiber 
pots. 

REFERENCES 

Frehner,  Ernst;  Fiirst,  Ernst.  1992.  Vom  Samen  bis  zur 
Pflanze.  Ein  Erfahrungsbericht  aus  dem  Forstgarten. 
Ber.  Eidgenoss.  Forsch.  anst.  Wald  Schnee  Landsch. 
333.  47  p. 

Mattes,  Hermann.  1982.  Die  Lebensgemeinschaft  von 
Tannenhaher  und  Arve.  Eidg.  Anst.  forstl. 
Versuchswes.,  Ber.  241,  74  S. 

Schonenberger,  Walter;  Frey,  Werner;  Leuenberger, 
Franz.  1990.  Oekologie  und  Technik  der  Aufforstung 
im  Gebirge  -  Anregungen  fiir  die  Praxis.  Eidg.  Anst. 
forstl.  Versuchswes.,  Ber.  325,  58  S. 

Translation:  Margaret  Sieber 


55 


DISTRIBUTION  AND  ECOLOGY  OF 
SIBERIAN  STONE  PINE  IN  THE  URALS 


P.  L.  Gorchakovsky 


Abstract — The  northern  and  southern  boundaries  of  the  distri- 
bution area  of  Siberian  stone  pine  (Pinus  sibirica)  in  the  Urals 
and  in  the  adjacent  plains  are  examined;  the  main  sociological 
factors  limiting  boundary  positions  are  described.  There  is  a 
natural  tendency  of  this  species  to  spread  northward  and  south- 
ward, from  places  recently  invaded,  but  this  process  is  usually 
stopped  or  turned  back  by  human  influence.  Numerous  occur- 
rences of  Siberian  stone  pine  beyond  its  present  continuous  range 
indicate  that  the  former  distribution  area  of  this  species  was 
larger  in  the  past  than  at  present. 


Siberian  stone  pine  (Pinus  sibirica),  or  briefly  Siberian 
pine,  is  considered  to  be  one  of  the  most  valuable  species 
of  coniferous  trees  growing  in  Russia.  Research  on  the 
limits  of  natural  distribution  of  Siberian  pine,  on  their 
causes  and  dynamics,  is  of  indisputable  theoretical  and 
practical  interest.  Research  is  especially  necessary  for 
scientific  substantiation  of  rational  use  of  all  forests,  par- 
ticularly those  with  Siberian  pine,  for  the  definition  of 
measures  to  protect  it  fi"om  being  exterminated,  and  for 
solving  the  problems  concerning  possible  cultivation  of 
this  species  in  adjacent  regions. 

The  general  distribution  area  of  Siberian  pine  includes: 
(1)  northeastern  regions  of  the  European  part  of  Russia 
(the  farthest  west  locality  is  in  the  upper  reaches  of  the 
Vychegda  River);  (2)  the  Ural  Mountain  range;  (3)  west- 
em  Siberia;  (4)  middle  and  eastern  Siberia  (up  to  the 
Aldan  River);  and  (5)  mountain  ranges  of  Altai,  the  Trans- 
baikal  area,  and  northern  Mongolia. 

The  main  limiting  factors  of  the  distribution  of  this 
species  are:  heat  deficiency  in  the  north;  permafi*ost,  peat- 
Ismd,  and  severe  winter  in  the  northeast;  water  deficiency 
and  low  air  hvunidity  in  the  south. 

The  western  part  of  the  area  is  located  in  the  Ural 
Mountains  and  in  the  northeastern  regions  of  the  Euro- 
pean part  of  Russia  adjacent  to  them.  Siberian  pine  occu- 
pies rather  large  areas  within  the  territory  of  Sverdlovsk 
and  Perm  provinces,  Komi  Republic,  and  within  Knanty- 
Mansy  national  district  of  Tyimien  province. 

THE  NORTHERN  LIMIT 

The  northern  border  of  the  distribution  area  (fig.  1) 
passes  along  the  Izhma-Pechora  watershed,  advances 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

P.  L.  Gorchakovsky  is  Professor,  Doctor  of  Biology,  Chief  Scientist  at 
the  Institute  of  Plant  and  Animal  Ecology,  Russian  Academy  of  Sciences, 
Ural  Branch,  620219,  Ekaterinburg,  Russia. 


considerably  northward  (up  to  65°30'  N.  latitude)  along 
the  valley  of  the  Pechora  River  and  even  farther  north- 
ward (67°  N.  latitude — the  northern  polar  circle).  Then 
it  switches  back  south  in  the  form  of  a  tongue,  to  the  foot- 
hills of  the  elevated  eastern  part  of  the  Russian  plain. 
From  Telpos-Iz  Mountain,  64°  N.  latitude,  it  leaps  north- 
ward along  the  eastern  slope  of  the  Poleir  and  Prepolar 
Urals  reaching  66°  10'  N.  latitude.  It  then  recedes  south- 
ward in  the  space  between  the  rivers,  but  returns  north- 
ward again  in  a  tongue-shaped  projection  along  the  vsdley 
of  the  Ob  River. 

Besides  that,  a  few  isolated  occurrences  of  Siberian  pine 
are  known  outside  its  main  distribution  area. 

The  northern  limit  of  Siberian  pine  does  not  reach  the 
extreme  boundaries  of  the  distribution  of  larch  (Larix 
sibirica)  and  spruce  (Picea  obovata).  The  well-known 
Russian  botanist  B.  N.  Gorodkov  (1929)  supposed  that  the 
advancement  of  Siberisin  pine  to  the  north  is  restricted  be- 
cause its  seeds  do  not  get  ripe  under  the  severe  conditions 
of  the  short  growing  season. 

However,  the  latest  observations  demonstrate  that 
even  at  the  northern  limit  of  the  distribution  area  seeds 
of  Siberian  pine  usually  get  ripe.  The  main  reason  likely 
is  that  Siberian  pine  is  not  able  to  develop  its  root  system 
imder  low  soil  temperature  (Tyrtikov  1954).  In  cold  soils, 
the  roots  of  Siberian  pine  ramificate  comparatively  weakly 
and  grow  more  slowly  than  the  roots  of  spruce  and  larch. 
Consequently,  insufficient  soil  heating  during  the  growing 
season  impedes  normal  development  of  root  systems,  and 
this  appears  to  be  the  main  factor  limiting  the  distribu- 
tion of  Siberian  pine  in  the  north.  Under  the  severe  cli- 
matic conditions  of  northern  Siberia,  pine  does  not  grow 
on  cold,  peatland  soils. 

Close  to  the  northern  boundaries,  in  the  upper  reaches 
of  the  Pechora  River,  Siberian  pine  selects  the  warmest 
well-heated  locations,  inhabiting  even  the  riverside  lime- 
stone outcrops  and  cliffs.  Contrary  to  that,  in  the  regions 
farther  south,  for  example  in  the  middle  and  southern 
Urals,  Siberian  pine  never  grows  in  such  habitats  because 
the  riverside  limestone  cliffs  are  too  warm  during  daylight 
periods. 

In  the  plains  adjacent  to  the  Urals  (the  Pechora  and 
western  Siberian  lowlands),  the  far-north  occurrences  of 
Siberian  pine  are  restricted  to  the  valleys  of  the  large  riv- 
ers (the  Pechora,  the  Ob,  and  some  of  their  tributaries). 
The  bovmdary  deviates  northward  in  the  form  of  narrow 
strips  along  the  river  valleys.  In  marshy  areas  between 
the  rivers,  however,  where  superficial  bedding  of  perma- 
fi'ost  fi*equently  occurs,  the  border  line  is  farther  south. 
Under  plain  conditions,  the  river  valleys  characteristically 
have  more  permeable  ground,  lower  levels  of  permafi"ost 
bedding,  and  warmer  soils  that  are  more  favorable  to 


56 


Figure  1 — Northern  distribution  limit  of  Siberian  stone  pine  in  tine  Urals  and  in  the  adjacent 
plains.  *  =  isolated  occurrences. 


Siberian  pine.  Moreover,  one  should  bear  in  mind  that 
the  valleys  of  the  large  rivers  that  bring  masses  of 
comparatively  warm  water  from  the  south  are  character- 
ized by  a  warmer  local  climate  favorable  to  the  growth  of 
Siberian  pine. 

At  the  Lyapin  depression,  in  the  region  of  Saranpaul, 
Siberian  pine- dominated  forests  are  to  be  found  on  buried 
palsa  peatbogs  covered  by  sandy  loam  deposits.  On  hill- 
ocks, better  conditions  for  drainage  and  heating  of  the  su- 
perficial layer  of  accumulated  deposits  favor  the  formation 
of  rather  tall  Siberian  pine  forests.  On  the  other  hand, 
raised  bogs  with  pools  and  ridges  between  the  hillocks  are 
treeless.  The  level  of  permafrost  here  is  rather  high;  ice- 
lenses  are  foimd  inside  the  peaty  hillocks. 

In  the  plains,  at  the  extreme  Hmit  of  its  distribution 
area,  Siberian  pine  is  usually  represented  by  solitary 
specimens  dispersed  within  spruce-larch-birch  forests; 
trees  usually  are  healthy,  without  signs  of  atrophy;  their 
tnmks  reach  10  to  14  m  in  height. 

Within  the  western  and  eastern  outskirts  of  the  Ural 
Moimtain  range,  the  Siberian  pine  border  displaces  north- 
ward along  meridionally  oriented  low  foothills  and  moim- 
tain ridges.  There,  this  species  grows  on  slopes  or  near 
brooks  and  small  rivulets  imder  moderate  or  slightly  in- 
creased (flowing)  ground  moisture  conditions.  The  advance 
northward  is  promoted  by  a  variety  of  sites  suitable  for  for- 
est growth  as  well  as  by  favorable  drainage,  better  soil 
heating,  and  consequently  by  lower  depth  of  permafrost. 

In  the  watershed  of  the  Urals  there  are  huge  mountains 
topped  by  woodless  tundrahke  "goltsy"  summits,  located 


within  the  mountain  tundra  and  high-mountain  cold 
desert  belts  (Gorchakovsky  1975,  1989).  They  are  char- 
acterized by  a  more  severe  climate,  and  consequently  the 
northern  limit  of  Siberian  pine  is  considerably  displaced 
southward. 

Siberian  pine  almost  absolutely  escapes  the  moimtain 
massif  of  the  Prepolar  Urals  from  the  west;  its  distribu- 
tion is  restricted  to  the  Pechora  lowland;  in  rare  cases  it 
occurs  in  the  moimtain  part  along  deep  erosion-tectonic 
valleys  of  the  large  rivers. 

On  the  eastern  slope  of  the  Prepolar  Urals,  however, 
Siberian  pine  ascends  rather  high  in  the  mountains,  al- 
most to  the  upper  forest  limit.  In  the  "subgoltsy"  belt 
(analogous  to  the  subalpine  belt),  in  the  region  of  Gorodkov 
Mountain  (Sale-Ur-Oika),  Siberian  pine  shows  dwarfed 
growth  (2.5  to  5  m  in  height),  exhibiting  trunks  thickened 
at  the  basal  portion  of  the  tree  and  tapering  upward,  and 
with  wide-branched  crowns  of  low  stability.  Frequently, 
such  dwarfed  trees  bear  abundant  cones.  At  lower  eleva- 
tions, within  the  mountain-taiga  belt  in  the  plain,  Sibe- 
rian pine  reaches  10  m  in  height.  From  the  eastern  side, 
along  the  foothills,  this  plant  extends  to  the  Polar  Urals 
reaching  66°  10'  N.  latitude  in  the  mountains  of  the  so- 
called  Small  Urals. 

More  intensive  advancement  of  Siberian  pine  north- 
ward, along  the  eastern  slope  of  the  Ural  Mountain  range, 
was  favored  partly  by  historical  causes.  This  species  in- 
vaded from  Siberia  to  new  places  in  the  Urals.  It  arrived 
earher  on  the  eastern  slope  of  the  Ural  Mountain  range, 
favored  mainly  by  the  specific  natural  conditions  of  the 
eastern  slope.  The  eastern  slope  of  the  Prepolar  and 


57 


Polar  Urals  has  a  more  continental  climate  than  the  west- 
ern slope.  Relative  air  humidity  is  lower  here;  annual 
precipitation  including  that  in  the  winter  is  less  heavy; 
and  snow  cover  melts  earlier  in  spring.  Consequently,  the 
growing  season  is  longer,  and  the  top  soil  layer  is  better 
heated. 

On  its  northern  boundaries,  Siberian  pine  is  distributed 
disjimctively  and  occurs  as  dispersed  small  clumps  or  soli- 
tary specimens  in  the  taiga.  All  these  extreme  localities 
are  rather  disjimct  from  each  other.  This  may  be  ex- 
plained primarily  by  the  fact  that  only  in  rare  locations 
are  the  combinations  of  various  environmental  factors 
favorable  for  growth  of  Siberian  pine  at  the  limits  of  its 
range.  On  the  other  hand,  the  dispersion  of  this  species 
has  been  reduced  here  and  there  due  to  forest  fires,  exten- 
sion of  bogs,  trampling  of  seedlings  and  young  growth 
by  reindeer,  and  extermination  by  humans.  In  locations 
where  human  activity  is  reduced  to  a  minimum,  Siberian 
pine  invades  new  places  north  of  its  present  limit.  This 
advance  is  observed  especially  clearly  in  the  mountain 
regions  of  the  Urals. 


THE  SOUTHERN  LIMIT 

The  southern  borderline  of  continuous  distribution  of 
Siberian  pine  (fig.  2)  passes  through  the  Russian  plain 
near  the  city  of  Perm;  then  it  turns  southward  along  the 
western  slope  of  the  middle  Urals,  reaching  the  most 
southern  position  at  56°  10'  N.  latitude;  then  it  switches 
back  north  following  the  eastern  slope  of  the  middle  Urals 
until  retiring  into  western  Siberia. 

A  few  outliers  occur  beyond  this  limit.  The  most  inter- 
esting of  them  are  to  be  found  in  the  mountain  dark  conif- 
erous taiga  of  the  southern  Urals  in  the  region  of  the  town 
of  Zlatoust  and  in  the  Ilmen  Mountains. 

At  the  southern  limit  of  its  range,  Siberian  pine  usually 
grows  in  the  form  of  single-standing,  tall  trees.  Only  on 
rocky  outcrops  does  it  exhibit  prostrated  growth  forms 
(krummholz). 

Natiiral  regeneration  of  Siberian  pine  near  the  southern 
boimdary  of  its  distribution  is  generally  satisfactory.  How- 
ever, there  are  many  factors  unfavorable  to  regeneration, 
such  as  seed  collection  difficulties,  fires,  and  trsimpling  of 


Figure  2— Southern  distribution  limit  of  Siberian  stone  pine  in  the  Urals 
and  in  the  adjacent  plains.  *  =  isolated  occurrences. 


58 


seedlings  by  livestock.  In  locations  where  such  vmfavor- 
able  influences  are  absent,  or  they  manifest  themselves 
to  a  lesser  degree,  regeneration  of  Siberian  pine  is  usually 
good.  So,  for  example,  Siberian  pine  reproduces  very  suc- 
cessfully on  raised  sites  protected  from  fire  within  the 
Bakhmet  Bog,  which  is  surroimded  by  Scots  pine  forests 
in  the  basin  of  the  Pyshma  River. 

Along  the  western  slope  of  the  Urals,  as  well  as  in  cer- 
tain sites  within  the  watershed  zone  where  the  climate  is 
more  humid,  Siberian  pine  has  invaded  new  places  signifi- 
cantly farther  south  than  in  the  Preural  and  Transural  re- 
gions. Its  southern  limit  projects  southward  in  the  form 
of  tongues  that  comprise  the  western  slope  and  part  of  the 
watersheds  of  the  moimtain  range.  Climate  of  the  eastern 
slope  of  the  middle  Urals  is  drier;  the  amoiint  of  precipita- 
tion is  less  compared  to  the  western  slope.  Therefore,  on 
the  eastern  slope  of  the  movmtain  range  and  in  the  foot- 
hills zone  adjacent  to  it  (the  region  of  predominance  of 
Scots  pine  forests),  Siberian  pine  occurs  considerably  far- 
ther north. 

Within  the  peatland  between  the  rivers  of  the  western 
Siberian  plain,  the  southern  limit  of  Siberian  pine  turns 
south  again.  There  Siberian  pine  grows  mainly  on  raised 
sites  within  bogs,  where  it  is  well  protected  from  fire. 

Siberian  pine  invaded  the  southern  Urals  later  than 
spruce  and  fir  {Abies  sibirica)  and  has  not  had  enough 
time  in  the  prehistoric  period  to  reach  the  climatically 
influenced  limit  of  its  distribution.  Human  activities 
hampered  the  southward  advance  of  Siberian  pine  into 
new  places  along  the  Ural  Mountain  range.  Here  and 
there,  numbers  of  this  species  have  declined.  Several  cen- 
turies ago  Siberian  pine  had  been  distributed  farther 
south  in  the  Urals  compared  to  its  contemporary  limit. 
Since  then  it  has  receded  northward  due  to  forest  fires 
and  himian  impact.  There  is  clear  evidence  of  recent  de- 
cline in  total  distribution.  In  the  southern  part  of  the 
area,  Siberian  pine  occurs  sporadically  in  small  groves, 
groups  of  trees,  or  as  single  trees.  It  grows  in  places  least 
endangered  by  fire,  most  frequently  on  moist  soils  in  the 
headwater  regions  and  valleys  of  brooks  and  rivulets,  on 
the  margins  of  bogs  and  on  raised  sites  within  them,  on 
lake  shores  (somewhere  close  to  water),  and  on  steep  out- 
crops of  acid  rocks. 

Formerly,  the  region  of  continuous  distribution  of  Sibe- 
rian pine  extended  south  at  least  to  the  isolated  localities 
mentioned  earlier,  but  probably  this  tree  species  extended 
even  farther  south.  Literature  data  as  well  as  oral  state- 
ments by  local  residents  provide  evidence  of  it.  Siberian 
pine  disappeared  from  these  places  not  only  due  to  cutting 
of  timber  and  forest  fires,  but  also  because  local  inhabit- 
ants picked  all  the  cones,  often  before  cones  were  com- 
pletely ripe.  They  also  dug  up  seedlings  for  transplanting 
into  personal  gardens. 

DISCUSSION  AND  CONCLUSIONS 

The  rather  complicated  picture  of  distribution  of  Sibe- 
rian pine  in  the  Urals  depends  to  a  large  extent  on  cli- 
matic difl'erences  between  the  relatively  himiid  western 
slope  of  the  range  and  the  more  continental  eastern  slope. 


Humidity  of  the  western  slope  restricted  the  northward 
advance  of  Siberian  pine,  but  favored  its  advance  to  the 
south.  On  the  contrary,  continentality  of  the  eastern 
slope  favored  the  advance  of  this  species  northward,  but 
restricted  its  advance  southward. 

At  both  the  northern  and  southern  Umits  of  its  distribu- 
tion, Siberian  pine  does  not  occur  high  up  in  the  moun- 
tains. Its  northernmost  occurrences  in  high  mountains 
are  in  the  eastern  part  of  the  Prepolar  Urals. 

In  the  northern  Urals,  Siberian  pine  is  widespread  al- 
most everywhere  in  the  high-mountain  belts.  On  certain 
mountains  (Oika-Nyor,  Pas-Nyor,  Chistop,  and  others), 
however,  it  forms  the  upper  forest  limit.  High-mountain 
Siberian  pine-dominated  forests  grow  on  steep  slopes  with 
poorly  developed  soils.  Near  the  upper  forest  limit,  such 
forests  may  be  foimd  mainly  in  less  high  mountains  far 
removed  from  large  mountain  massives.  In  such  isolated 
mountains,  the  upper  forest  limit  is  caused  chiefly  by 
edaphic  factors,  while  the  cHmate  of  their  treeless  zone 
and  of  the  subgoltsy  belt  is  less  severe.  Dwarfed,  pros- 
trate, and,  in  most  cases,  noncone-producing  specimens 
of  Siberian  pine  of  the  northern  Urals  occur  in  high- 
moimtain  tundras  significantly  higher  in  elevation  than 
the  normal  upper  forest  limit.  For  instance,  dwarf  speci- 
mens occur  on  the  movmtains  Isherim,  Oika-Chahl, 
Yalping-Nyor,  Denezhkin  Kamen,  Konzhakovsky  Kamen, 
Kosvinsky  Kamen,  and  others.  The  occurrences  of  kmmm- 
holz  growth  forms  above  the  normal  upper  forest  limit 
have  to  be  explained  by  the  dispersal  of  seeds  into  the 
moimtain  tundra  zone  by  the  nutcracker  and  rodents. 

In  the  middle  Urals  near  the  simamits  of  relatively  high 
moimtains,  Siberian  pine  also  forms  singly  standing, 
dwarfed  trees  (the  mountains  Kachkanar,  Starick-Kamen, 
and  others).  South  of  Moimt  Starick-Kamen,  this  species 
is  never  foimd  in  high-movmtain  belts,  but  in  the  forest 
belt  its  area  spreads  considerably  farther  south. 

Judging  by  its  biological  and  ecological  features,  Sibe- 
rian pine  may  not  be  considered  as  a  species  becoming 
extinct.  It  is  perfectly  well  adapted  to  the  taiga  environ- 
ment; it  is  a  good  cone  producer,  and,  if  human  impact 
does  not  interfere,  it  regenerates  quite  satisfactorily.  Al- 
most everywhere,  regeneration  occurs  close  to  old  trees. 
However,  one  should  bear  in  mind  that  Siberian  pine 
grows  more  slowly  than  other  shade-tolerant  coniferous 
trees.  Therefore,  ground  fires  that  annihilate  second- 
growth  trees  adversely  afl'ect  natural  regeneration  of  this 
species. 

The  border  lines  of  Siberian  pine,  especially  on  the 
southern  border,  have  been  reduced  by  careless  and  some- 
times barbarian  treatment  of  this  valuable  plant.  It  is 
necessary  to  secure  proper  protective  measiires  for  its 
southern  localities  in  the  Urals  and  to  use  seeds  gathered 
from  trees  mainly  for  sowing  under  the  forest  canopy  and 
in  tree  nurseries. 

In  the  coTorse  of  intensive  harvest  cuttings  in  the  dark 
coniferous  fir-spruce  forests  that  have  individual  Siberian 
pine  trees  in  the  admixture,  foresters  usually  retain  Sibe- 
rian pine  as  seed  trees  for  natural  regeneration.  How- 
ever, in  many  cases,  this  does  not  work  well  because  in 
the  heavy  clay  and  loamy  soils  of  the  plains,  and  in  the 


59 


mountains  on  poorly  developed  soils,  such  trees  r\in  the 
danger  of  windfall.  To  avoid  windfall,  it  is  necessary  to 
retain  clumps  of  Siberian  pine  with  accompanying  tree 
species  instead  of  single  specimens  of  Siberian  pine. 

Natural  conditions  of  the  middle  Urals  and  of  the 
moiintain-range  portion  of  the  southern  Urals  (at  least 
up  to  54°  N.  latitude)  are  favorable  for  regeneration  and 
growth  of  Siberian  pine.  This  valuable  tree  species  may 
be  reintroduced  to  areas  south  of  its  present  distribution. 
This  can  be  concluded  from  the  successful  cultivation  of 
Siberian  pine  south  of  its  present  border  line  in  the  moun- 
tain taiga  regions  of  the  middle  and  southern  Urals. 


REFERENCES 

Gorchakovsky,  P.  L.  1975.  The  plant  world  of  the  high- 
mountain  Urals.  Moscow:  "Nauka"  Publishing.  284  p. 
[In  Russian]. 

Gorchakovsky,  P.  L.  1989.  Horizontal  and  altitudinal 
differentiation  of  the  vegetational  cover  of  the  Ural 
Mountains.  Pirineos,  N  133:  33-54. 

Gorodkow,  B.  N.  1929.  The  Polar  Urals  in  the  upper 
reaches  of  the  rivers  Voykar,  Synya  and  Lyapin.  In: 
Materials  of  the  Committee  for  Expeditional  Re- 
searches, V.  7,  Leningrad:  1-28.  [In  Russian]. 

Tyrtikov,  A.  P.  Growth  of  tree  roots  at  the  northern  limit 
of  forests.  Bull,  of  Moscow  Soc.  of  Naturalists.  Section 
of  Biology.  59(1):  72-82.  [In  Russian]. 


60 


ECOPHYSIOLOGICAL  INVESTIGATIONS 
ON  CEMBRAN  PINE  AT  TIMBERLINE  IN 
THE  ALPS,  AN  OVERVIEW 

Rudolf  Hasler 


Abstract — During  the  nineteenth  century  foresters  as  well  as 
scientists  became  interested  in  learning  more  about  cembran 
pine  (Pinus  cembra)  in  the  Alps,  but  extensive  ecophysiological 
fieldwork  did  not  begin  before  the  1930's.  Great  progress  was 
made  after  the  estabUshment  of  permanent  field  stations  at  the 
timberUne  at  Obergurgl  (1953),  Patscherkofel  near  Innsbruck 
(1963),  and  Stillberg  near  Davos  (1959).  In  a  brief  overview, 
some  selected  results  of  ecophysiological  research  in  cembran 
pine  are  presented. 


The  old  and  bizarre  cembran  pines  (Pinus  cembra  L.), 
which  form  the  timbedine  over  a  wide  range  in  the  Cen- 
tral Alps,  have  always  fascinated  himaans.  In  earlier 
days,  the  tree  was  of  some  importance.  It  produced  not 
only  wood  but  also  nuts  that  could  be  used  as  food.  There- 
fore, this  tree  has  been  of  some  interest  for  a  long  time, 
and  a  lot  of  books  and  papers  about  cembran  pine  in  the 
Alps  have  been  published,  in  forestry  as  well  as  in  biologi- 
cal science. 

This  short  overview  presents  only  some  selected  ex- 
amples of  the  ecophysiological  research,  with  a  special 
focus  on  gas  exchange.  A  general  overview  on  ecophysio- 
logical problems  at  the  timberline  was  published  by 
TranquilUni  (1979).  Many  other  subjects,  such  as  growth, 
genetics,  and  mycorrhiza,  have  been  investigated  in 
Austria,  France,  Italy,  and  Switzerland  but  are  not  men- 
tioned in  this  overview. 

BEGINNING  OF  INVESTIGATIONS 

In  the  middle  of  the  nineteenth  century,  foresters  be- 
came interested  in  the  cembran  pines  growing  in  the  Alps. 
Fankhauser  (1853)  described  the  distribution  of  these 
trees  together  with  the  climatological  conditions  they 
need  for  growing  well.  He  mentioned  that  it  is  not 
possible  to  transfer  this  tree  species  to  low  elevations. 
Kemer  (1866)  and  Simony  (1870)  pubhshed  similar  data 
for  parts  of  Austria.  At  the  same  time,  the  diseases  and 
the  insect  enemies  of  the  cembran  pine  in  the  Alps  at- 
tained some  importance.  They  have  been  described  by 
Keller  (1890)  and  others. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Rudolf  Hasler  is  Head  of  the  Section  "Forest  and  Climate,"  Swiss  Fed- 
eral Institute  for  Forest,  Snow  and  Landscape  Research,  CH-8903 
Birmensdorf,  Switzerland. 


ECOLOGICAL  AND  GEOGRAPfflCAL 
OBSERVATIONS 

More  and  more  scientists  became  interested  in  cembran 
pine  at  the  end  of  the  nineteenth  century  and  the  early 
twentieth  century  to  learn  more  about  these  trees,  which 
grow  on  the  most  extreme  sites  in  the  mountains.  Mono- 
graphs concerning  cembran  pine  in  Switzerland  (RikH 
1909)  and  in  Austria  (Nevole  1914)  were  pubhshed.  Many 
other  publications  were  written  at  this  time.  The  main 
subjects  were  growth,  and  again  fungal  diseases  and  prob- 
lems with  insects  (Fankhauser  1903;  Keller  1901,  1910; 
and  others). 

FIRST  ECOPHYSIOLOGICAL  FIELD 
EXPERIMENTS 

Ecophysiological  investigation  in  the  field  began  in  the 
thirties  on  Patscherkofel  near  Innsbruck  (Austria).  To 
work  near  the  timberline  in  the  Alps  was  a  logistic  prob- 
lem. The  cembran  pines  only  grow  near  the  timberline, 
at  high  altitudes.  It  was  extremely  hard  to  work  there  in 
earher  times. 

Cartellieri  (1935)  was  probably  the  first  to  measiu*e 
photosynthesis  and  water  relations  throughout  a  whole 
season  in  cembran  pine  at  timberline.  At  the  same  time 
he  also  took  measurements  of  different  dwarf  shrubs  at 
high  elevation.  He  measured  photosynthesis  by  absorbing 
the  carbon  dioxide  in  barite-water,  and  then  he  deter- 
mined the  COg  assimilation  by  titration  in  the  laboratory. 
Some  idea  about  the  difficulties  of  working  at  the  timber- 
line  in  those  days  is  given  in  his  papers:  missing  data  in 
the  annual  course  are  explained  as  "walking  back  to  the 
laboratory  I  lost  some  of  the  probes  because  some  of  the 
glass  flasks  were  broken,"  or  in  another  example,  "at  this 
time  I  got  ill  and  had  to  stop  the  investigation."  Never- 
theless his  results  show  more  or  less  the  same  values  as 
we  are  measuring  today  with  ovu*  modem  climatized  and 
very  expensive  instrumentation.  Also  about  the  same 
time  Ulmer  (1937)  investigated  the  annual  course  of  fi*ost 
hardiness  and  measured  in  parallel  the  osmotic  values 
in  the  needles. 

PERMANENT  FIELD  STATIONS  AND 
CONTINUOUS  CO2  MEASUREMENTS 

Great  progress  in  timberline  research  in  the  Alps  was 
made  in  the  1950's.  Again  in  Austria,  a  permanent 
experimental  field  station  at  an  altitude  of  about  1,950  m 
above  sea  level  (a.s.l.)  was  established  in  1953  near 


61 


Obergurgl.  At  the  same  time,  the  newly  developed  infra- 
red gas  analyzer  (IRGA)  made  it  possible  to  measm'e 
continuously  carbon  dioxide  in  the  air. 

Tranquillini  (1955)  conducted  experiments  at  the  new 
field  station  at  Obergurgl  to  investigate  the  relations  be- 
tween light,  temperature,  and  photosynthesis  in  needles 
of  young  cembran  pines  growing  in  the  sun  or  in  the  shad- 
ows. He  determined  the  light  compensation  points  for 
these  needles  as  500  and  150  Lux,  respectively,  and  the 
light  saturation  points  as  25  and  18  kLux,  respectively. 
The  COg  assimilation  values  he  reported  reached  2.5  mg 
and  2.1  mg  COJg  dry  weight,  respectively.  Net  photo- 
synthesis starts  at  -5  °C  and  has  an  optimum  range  be- 
tween 10  and  15  °C.  The  upper  temperature  limit  is  at 
35  °C. 

The  microclimatic  conditions  during  wintertime  and 
needle  and  soil  temperatures  as  well  as  gas  exchange  and 
water  relations  in  young  cembran  pines  growing  on  natu- 
ral sites  at  the  timberline  were  published  by  Tranquillini 
(1957).  He  showed  that  in  wintertime,  even  in  warmer 
weather  conditions,  no  net  photosynthesis  was  detectable. 
Holzer  (1958)  explained  this  lack  of  photosynthesis  by  a 
changed  cell  structure  as  well  as  by  a  changed  physiolo- 
gical behavior  during  wintertime. 

Pisek  and  Winkler  (1958)  measured  the  annual  course 
of  the  capacity  of  net  photosynthesis  and  of  respiration  in 
detached  shoots  of  cembran  pines  growing  at  different  al- 
titudes. During  wintertime,  they  found  an  always  nega- 
tive carbon  dioxide  balance  in  the  trees  at  the  timberline, 
whereas  those  growing  at  lower  elevations  could  produce 
a  positive  one,  at  least  during  periods  with  temperatures 
above  the  freezing  point.  Keller  (1970)  confirmed  these 
findings,  using  another  method.  Tranquillini  (1958, 
1963d)  investigated  the  annual  course  of  frost  resistance. 
He  found  that  young  trees  are  more  sensitive  to  frost  than 
mature  trees.  Young  trees  become  damaged  by  tempera- 
tures between  -15  to  -30  °C,  while  mature  trees  can  sur- 
vive temperatures  lower  than  —45  °C.  Keller  and  Beda- 
Puta  (1973)  analyzed  stem  respiration  in  cembran  pine 
during  wintertime. 

Pisek  and  Winkler  (1959)  studied  photosynthesis  in  re- 
lation to  temperature  in  different  light  conditions  as  well 
as  at  different  altitudes.  With  increasing  light  intensity, 
the  temperature  optimum  is  shifted  to  higher  tempera- 
tures. There  are  some  difficulties  in  comparing  all  these 
older  results  with  the  measurements  of  today,  because  not 
all  parameters  can  be  transformed  (for  example,  illumina- 
tion [Lux]).  Some  other  values,  such  as  the  needle  sur- 
faces, were  normally  not  used  as  a  reference  for  gas  ex- 
change measurements  before  the  1980's. 

A  COg  budget  for  a  whole  year  was  estimated  by 
TranquiUini  (1959a,  1959b,  1963a,  1963b).  Total  CO^ 
uptake  per  year  (gross  photosynthesis)  of  a  5-  to  8-year- 
old  tree  at  2,000  m  a.s.l.  was  about  7.8  kg.  About  38  per- 
cent was  lost  by  respiration  from  needles,  branches,  stem, 
and  roots.  According  to  the  measured  net  COg  assimi- 
lation, a  tree  should  produce  2.2  g  dry  matter  per  gram  of 
needles.  In  reality  they  produced  0.65  g.  The  author 
mentioned  that  most  of  the  missing  carbon  dioxide  may 
have  been  used  by  the  mycorrhiza. 

In  the  second  half  of  the  1950's,  ecological  investiga- 
tions with  the  aim  of  finding  methods  for  successful 


afforestation  at  high  altitude  were  started  in  Switzerland 
(Stillberg,  Davos:  2,000-2,230  m  a.s.l.)  and  in  Austria 
(Haggen,  Sellraintal:  1,730-1,900  m  a.s.l.).  Cembran  pine 
was  one  of  the  species  used  in  these  field  experiments, 
which  still  are  going  on.  It  takes  a  long  time  to  get  results 
from  trees  growing  at  high  elevation  near  the  timberline. 
Svmimaries  containing  prelimin£iry  results  were  published 
by  Schonenberger  and  Frey  (1988)  for  Stillberg  and  by 
Kronfuss  (1980,  1983,  1986)  for  Haggen. 

LABORATORY  CLIMATE  CONTROL 

The  technical  possibilities  developed  fast.  On 
Patscherkofel  (1,950  m  a.s.l.)  a  climate  laboratory  was 
established  in  1963.  In  this  laboratory  a  device  was  in- 
stalled that  allows  analysis  of  the  gas  exchange  of  yoimg 
trees  in  climatized  windchannels  at  high  altitude.  Now  it 
became  possible  to  control  exactly  the  climatic  parameters 
such  as  temperature  and  air  and  soil  humidity  as  well  as 
windspeed,  which  was  a  brand-new  innovation.  Using 
this  new  instrumentation  old  measurements  could  be  ve- 
rified and  a  lot  of  new  possibilities  were  opened  up.  In 
particular,  water  relations  could  now  be  investigated  un- 
der controlled  climatic  conditions. 

Tranquillini  (1963c)  investigated  the  CO^  assimilation 
in  relation  to  air  and  soil  humidity.  In  water-saturated 
soil,  dry  air  (25  percent  relative  hmnidity)  reduced  photo- 
synthesis in  cembran  pines  by  about  33  percent.  A  little 
less  water  in  the  soil  (half  of  field  capacity)  allowed  some 
higher  photosjnithetic  rates.  Resistance  to  dry  himiidity 
increases  from  spring  to  autimin  (Tranquillini  1965). 

Pisek  and  others  (1967,  1968, 1969)  explored  the  tem- 
perature limits  of  photosynthesis  to  chill  and  frost,  and  to 
high  temperatures.  COg  assimilation  is  operative  between 
-4.7  and  +36  °C,  but  the  trees  can  survive  temperatures 
between  below  -50  and  up  to  48  °C.  Of  great  importance 
is  not  only  the  time  factor,  but  also  the  physiological  con- 
dition of  the  plant.  Temperature  hardiness  changes  with 
the  seasons,  £md  young  trees  or  seedlings  are  more  sensi- 
tive to  extreme  temperatures  than  mature  trees.  The  re- 
lation between  photosynthetical  capacity  and  temperature 
was  analyzed  by  Tranquillini  and  Machl-Ebner  (1971). 
Another  factor  investigated  was  day  length.  Bamberg 
and  others  (1967)  and  Schwarz  (1970a,  1970b,  1971)  have 
shown  that  day  length  does  not  influence  the  photosyn- 
thetic  capacity  during  winter  depression,  but  trees  kept 
in  warm  conditions  during  winter  commenced  growth  in 
spring  much  earlier  than  plants  in  nature.  The  variation 
in  frost  resistance  of  the  trees  changed  in  parallel  with 
the  temperature.  In  long-day  regimes,  frost  resistance  of 
cembran  pine  remained  at  a  lower  level  than  those  grow- 
ing imder  short-day  conditions.  The  researchers  supposed 
that  an  endogenous  rhythm  also  influences  resistance  to 
extreme  temperatures. 

Caldwell  (1970),  using  the  climatized  wind  tunnel  of  the 
Patscherkofel  laboratory,  has  shown  the  effect  of  wind  on 
stomatal  aperture,  photosynthesis,  and  transpiration. 
He  did  not  detect  a  great  influence  of  low  windspeeds  on 
these  parameters.  At  higher  windspeeds  (up  to  8  m/sec) 
photosynthesis  and  transpiration  were  slightly  reduced. 
Baig  and  others  (1974)  and  Baig  and  Tranquillini  (1980) 
investigated  the  cuticular  transpiration  rate  of  cembran 


62 


pine  and  found  that  cuticular  transpiration  in  this  tree 
is  very  slow  but,  as  in  other  species,  cuticular  resistance 
decreases  in  paredlel  to  the  elevation  at  which  the  shoots 
were  collected. 

Havranek  (1972)  found  that  cold  soil  temperatures 
reduce  photosynthesis  and  transpiration  in  cembran  pine. 
Havranek  and  Benecke  (1978)  investigated  the  influence 
of  soil  moistiire  on  transpiration  and  photosynthesis. 
Cembran  pine  begins  to  close  the  stomata  at  a  soil  water 
potential  of  about  -0.4  bars.  Compared  to  other  species, 
cembran  pine  maintains  a  high  water-use  efficiency  and 
uses  limited  soil  water  economically.  Giinthardt  and 
Wanner  (1982)  showed  that  needles  from  timberline  trees 
have  on  the  average  more  wax  than  those  grown  at  a 
lower  elevation. 


Another  step  to  get  more  precise  resvilts  for  the  CO^ 
balance  of  cembran  pine  was  taken  by  Tranquillini  and 
Schiitz  (1970).  They  measured  bark  respiration  at  the 
timberline  and  reported  that  b£irk  respiration  also  shows 
an  annual  course.  In  general,  bark  respiration  of  cem- 
bran pine  was  rather  high  compared  to  that  of  other  spe- 
cies. The  loss  of  carbon  dioxide  by  stem  respiration 
throughout  a  year  was  calculated  to  be  about  18.5  percent 
of  gross  photosynthesis.  Measuring  stem  respiration  con- 
tinuously, Havranek  (1981)  found  that  there  is  a  marked 
fluctuation  in  stem  respiration  during  the  growing  season, 
but  he  could  not  find  any  relation  between  activity  of  stem 
respiration  and  tree  ring  growth. 


Net  photosynthesis 

40- 

■a  32" 
E  2411 

o  16 
E  81 
E 

-8 


I  I  I  I   I  B  I  I  I   I   I  11  I  I  I   I  I  11  I  I  I  I  I  11  I  I  I  I  I   11  I  I   I   I  I  I  I   I   I  I  I  11  I   I   I  I  I  H  I  I  I  I  I  11  I  I  I 

5  15  2515  15  2515  15  2515  15  25  15  15  25  1 5  15  251  5  15  2515  15  2515  15  25  I  5  15 
Photosynthetically  active  radiation 


'l  I  I  I  I  I  11  I  I  I  I  I  11  rr  I  I  I  11  I  I'V  ri  iTTi  i  i  i  ii  T  i  i  i  i  i  i  i  i  i  i-ir  i  i  i  i  i  n  i  i  i  i  i  ii  i  i  r 
15  15  2515  15  2515  15  251  5  15  25  I  5  15  25  I  5  15  2515  15  25  I  5  15  2515  15  25  I  5  15 

Air  temperature 


5  15  2515  15  2515  15  251  5  15  25  I  5  15  25  1 5  15  2515  15  2515  15  251  5  15  25  I  5  15 


Soil  temperature 


I  M  III  I  I  I  I  II  I  I  I  I  I  II  I  I  I  I  I  11  I  I  I  I  I  11  I  I  I  I  I  I  I  (  I  I  I  11  I  I  I  I  I  11  I  I  I  I  I  11  I  I  !' 

1 5  15  2515  15  25  1 5  15  2515  15  25  1 5  15  2515  15  251  5  15  25  I  5  15  2515  15  2515  15 


Water  vapor  pressure  difference 


Is  1525'! 


15  15  2515  15STr'^l525 


Sep  Oct 


Nov 


5  15  25  15  15  2515  15  251 

Dec     Jan      Feb  Mar 


5  15 

Apr     May  Jun 


25 
20 
15 
10 
5 
0 
-5 
-10 
-15 
-20 


Figure  1 — Photosynthesis  in  cembran  pine, 
photosynthetically  active  radiation,  air  and  soil 
termperature,  and  air  humidity  measured  at  the 
timberline  (Stillberg,  1,980  m  a.s.l.,  1979-80). 


63 


FIELD  CLIMATE  CONTROL 

In  the  late  sixties,  the  thermoelectrically  climatized  gas 
exchange  chambers  were  developed  by  Koch  and  others 
(1968).  Hasler  and  Blaser  (1981)  measured  the  first  daily 
courses  in  young  cembran  pines  on  the  east-  and  the 
north-facing  slopes  of  an  avalanche  gully  at  the  timberline 
at  Stillberg.  The  two  sites  exemplify  marked  differences 
in  microclimate  (mean  values  of  soil  surface  temperatures 
per  growing  season  differ  by  more  than  10  °C).  These 
differences  in  microclimate  also  £iffect  the  photosynthetic 
rates  of  afforested  cembran  pines.  Carbon  dioxide  gain, 
summarized  over  a  season,  of  the  trees  on  the  north- 
facing  slope  was  34  percent  less  them  for  those  on  the 
east-facing  slope  (Turner  and  Streule  1983). 

Gas  exchange  measured  in  combination  with  meteoro- 
logical parameters  was  registered  in  a  branch  of  an  older 
cembran  pine  at  Stillberg  (1,980  m  a.s.l.)  throughout 
a  whole  winter  season  in  1979-80  (fig.  1).  Havranek 
(1981)  published  an  annual  course  of  growth  and  a 
carbon  dioxide  balance,  taken  near  the  Patscherkofel 
laboratory.  During  summer  the  measurements  were 
taken  with  climatized  chambers. 

In  the  middle  of  the  1980's  the  so  called  "forest  decline" 
had  high  priority  in  forest  research  in  Europe.  The  main 
work  in  this  field  was  done  on  other  tree  species  than 
cembran  pine.  Nevertheless,  some  investigations  were 
made  on  this  species.  Havranek  (1987)  took  measure- 
ments on  cembran  pine  at  the  timberline  to  investigate 
stress  caused  by  air  pollution  in  combination  with  climatic 
factors.  Genys  and  Heggestad  (1978)  did  not  find  a  high 
susceptibility  of  cembran  pines  to  ozone  and  sulphur  di- 
oxide, but  MinarCic  and  KubiCek  (1991)  suggested  that 
dust  could  influence  the  stomata  in  needles  of  cembran 
pines  and  thus  affect  photosynthesis.  Lutz  and  others 
(1988)  emalyzed  the  photos)nithetic  pigments  of  cembran 
pines  growing  at  different  altitudes,  which  thus  are  ex- 
posed to  different  ozone  concentrations.  Nebel  and  Matile 
(1992)  investigated  the  cause  of  uncommon  yellowing  of 
the  needles  of  cembran  pines  by  analyzing  photosynthe- 
sis, nutrient  relations,  longevity,  and  senescence  of  these 
declining  needles. 

The  newest  measm*ements  in  cembran  pine  from  the 
Stillberg  experimental  station  are  presented  in  these  pro- 
ceedings (Koike  and  others).  These  are  probably  the  first 
measvirements  of  photosynthesis  in  cembran  pine  at  the 
timberline  with  an  eirtificially  elevated  atmospheric 
concentration.  A  detailed  publication  giving  these  results 
is  in  preparation  (Koike  and  others  1993). 

CONCLUSIONS 

At  the  beginning  of  the  research  on  cembran  pine,  about 
150  years  ago,  scientists  worked  mainly  at  the  forest 
level.  With  the  development  of  sophisticated  measure- 
ment instruments,  branches,  twigs,  or  even  needles  were 
the  focus  of  interest.  Rarely  did  anyone  try  to  extrapolate 
these  results  from  the  small  parts  to  a  whole  tree,  a  prob- 
lem which  has  still  not  been  satisfactorily  resolved.  Prob- 
ably we  should  try  to  go  back  directly  to  the  level  of  whole 
trees  or  even  of  stands,  especially  when  we  want  to  under- 
stand changes  caused  by  external  influences,  such  as  air 


pollution,  rising  carbon  dioxide,  changing  UV  radiation, 
and  many  others. 

ACKNOWLEDGMENT 

I  thank  Mrs.  M.  J.  Sieber  for  critical  editing  of  the 
English  version  of  my  manuscript. 

REFERENCES 

Baig,  M.  N.;  Tranquillini,  W.  1980.  The  effect  of  wind  and 
temperature  on  cuticular  transpiration  of  Picea  abies 
and  Pinus  cembra  and  their  significance  in  desiccation 
damage  at  alpine  treeline.  Oecologia.  47:  252-256. 

Baig,  M.  N.;  Tranquillini,  W.;  Havranek,  W.  M.  1974. 
Cuticulare  Transpiration  von  Picea  abies-  und  Pinus 
cembra-Zvfeigen  aus  verschiedener  Seehohe  imd  ihre 
Bedeutung  fiir  die  winterliche  Austrocknung  der 
Baiune  an  der  alpinen  Waldgrenze.  Centredblatt  fiir 
das  gesamte  Forstwesen.  91:  195-211. 

Bamberg,  S.;  Schwarz,  W.;  TranquiUini,  W.  1967.  Influen- 
ce of  daylength  on  photosynthetic  capacity  of  stone  pine 
(Pinus  cembra  L.).  Ecology.  48:  264-269. 

Caldwell,  M.  M.  1970.  The  effect  of  wind  on  stomatal 
aperture,  photosynthesis,  and  transpiration  of  Rhodo- 
dendron ferrugineum  L.  and  Pinus  cembran  L. 
Centralblatt  fur  das  gesamte  Forstwesen.  87:  193-201. 

Cartellieri,  E.  1935.  Jahresgang  von  osmotischem  Wert, 
Transpiration  und  Assimilation  einiger  Ericaceen  der 
alpinen  Zwergstrauchheide  und  von  Pinus  cembran. 
Jahrbiicher  fiir  wissenschaftliche  Boteuiik.  82:  460-506. 

Fankhauser,  F.  1853.  Uber  das  Vorkommen  und 
Gedeihen  der  Arve.  Schweizerisches  Forst  Journal. 
S83. 

Fankhauser,  F.  1903.  Der  Kiefemschiittepilz  an  der  Arve. 
Schweizerische  Zeitschrifl  fiir  Forstwesen.  54:  321-323. 

Genys,  J.  B.;  Heggestad,  H.  E.  1978.  Susceptibility  of  dif- 
ferent species,  clones  and  strains  of  pines  to  acute  in- 
jury caused  by  ozone  and  sulfur  dioxide.  Plant  Disease 
Reporter.  62(8):  687-691. 

Gtinthardt,  M.  S.  ;  Wanner,  H.  1982.  Die  Menge  des 
cuticularen  Wachses  auf  Nadeln  von  Pinus  cembran 
(L.)  und  Picea  abies  (L.)  Karsten  in  Abhangigkeit  von 
Nadelalter  und  Standort.  Flora.  172:  125-137. 

Hasler,  R.;  Blaser,  P.  1981.  Nettophotosjmtheseaktivitat 
von  Arve  imd  Larche  auf  verschiedenen  Standorten  der 
subalpinen  Stufe.  Mitteilungen  der  forstlichen 
Bundesversuchsanstalt,  Wien.  140:  179-188. 

Havranek,  W.  1972.  The  influence  of  soil  temperature  on 
photosynthesis  and  transpiration  of  young  trees  and  on 
dry-matter  production,  at  the  timber  line.  Angewandte 
Botanik.  46:  101-116. 

Havranek,  W.  M.  1981.  Stammatmung,  Dickenwachstum 
und  Photosynthese  einer  Zirbe  {Pinus  cembra)  an  der 
Waldgrenze.  Mitteilimgen  der  forstlichen  Bundes- 
versuchsanstalt, Wien.  142:  443-467. 

Havranek,  W.  M.  1987.  Physiologische  Reaktionen  auf 
Klimastress  bei  Baumen  an  der  Waldgrenze.  Sympo- 
smm  Klima  und  Witterung  in  ZusEunmenhang  mit  den 
neuartigen  Waldschaden.  Tagungsbeitrage,  Miinchen: 
115-138. 


64 


Havranek,  W.  M.;  Benecke,  U.  1978.  The  influence  of  soil 
moistiire  on  water  potential,  transpiration  and  photo- 
synthesis of  conifer  seedlings.  Plant  and  Soil.  49(1): 
91-103. 

Holzer,  K.  1958.  Die  winterlichen  Veranderungen  der 
Assimilationszellen  von  Zirbe  (Pinus  cembra  L.  )  und 
Fichte  (Picea  excelsa  Link.  )  an  der  alpinen  Waldgrenze. 
osterreichische  botanische  Zeitschrift.  105:  323-346. 

Keller,  C.  1890.  Tierische  Forstbeschadigiingen  an  der 
Arve.  Osterreichische  Forstzeitung:  267-268. 

Keller,  C.  1901.  Die  Arven-Erkrankungen  im 
Oberengadin.  Schweizerische  Zeitschrift  fur  Forst- 
wesen.  52:  293-299. 

Keller,  C.  1910.  Die  tierischen  Feinde  der  Arve. 
Schweizerische  Centralanstalt  fur  forstliches.  Versuchs- 
wesen,  Mitteilungen.  10(1):  1-50. 

Keller,  T.  1970.  Uber  die  Assimilation  einer  jungen  Arve 
in  Winterhalbjahr.  Biindnerwald.  23(2):  49-54. 

Keller,  T.;  Beda-Puta,  H.  1973.  Gas  exchange  of  leafless 
shoots  in  winter.  Schweizerische  Zeitschrift;  fiir 
Forstwesen.  124(6):  433-441. 

Kemer,  A.  1866.  Wachstumsbedingungen  der  Zirbe. 
osterr.  Monatsschr.  Forstwesen.  1-15. 

Koch,  W.;  Klein,  E.;  Walz,  H.  1968.  Neuartige 
Gaswechsel-Messanlage  fiir  Pflanzen  in  Laboratorium 
und  Freiland.  Siemens  Zeitschrift;.  42:  392-404. 

Koike,  T.;  Hasler,  R. ;  Matyssek,  R.;  Item,  H.  1993.  Sea- 
sonal changes  in  the  photosynthetic  capacity  of  Larix 
decidua  and  Pinus  cembran  planted  on  contrasting 
slopes  at  the  timberline  at  Stillberg  (Davos),  eastern 
Switzerland.  [In  preparation]. 

Kronfiiss,  H.  1980.  Das  Bestandesklima  einer 
HochlagenaufForstung  auf  einem  Sonnenhang.  In: 
Beitrage  zur  subalpinen  Waldforschung.  2.  Folge. 
Mitteilvmgen  der  forstlichen  Bundesversuchsanstalt, 
Wien.  129:  81-103. 

Kronfuss,  H.  1983.  Das  Hohenwachstum  einer  Zirben- 
aufforstung  auf  einem  Siidhang  in  klimatologischer 
Sicht.  Allgemeine  Forstzeitung.  94(12):  330-332. 

Kronfuss,  H.  1986.  Die  Abhangigkeit  des  Hohen- 
zuwachses  der  Zirbe  von  der  Seehohe.  Jahresbericht 
1985.  Forstliche  Bundesversuchsanstalt,  Wien:  137-144. 

Lutz,  C;  Schulte-Hostede,  S.;  Kirchner,  M.;  Reuther  M. 
1988.  Photosjnithetische  Pigmente  aus  Nadelbaumen 
unterschiedlicher  Hohenstufen  des  Otztals.  Internatio- 
nales Symposium  ''Verteilung  und  Wirkimg  von  Photo- 
oxidantien  im  Alpenraum.  Sitzimgsberichte  Miinchen 
1988:  415-425. 

MinarCic,  P.;  KubfCek,  F.  1991.  Localization  of  emissions 
on  the  needle  surface.  1.  Methodology,  morphological 
characterization  of  the  stomata  {Picea  abies,  Pinus 
cembran).  Ekologia,  CSFR.  10(4):  405-413. 

Nebel,  B.;  Matile,  Ph.  1992.  Longevity  and  senescence 
of  needles  in  Pinus  cembran.  Trees.  6(3):  156-161. 

Nevole,  J.  1914.  Die  Verbreitimg  der  Zirbe  in  der  osterr.  - 
ungar.  Monarchic.  Verlag  W.  Frick  Wien.  89  p. 

Pisek,  A.;  Winkler,  E.  1958.  Assimilationsvermogen  und 
Respiration  der  Fichte  (Picea  excelsa  Lind. )  in  verschie- 
dener  Hohenlage  und  der  Zirbe  (Pinus  cembran  L.  )  an 
der  alpinen  Waldgrenze.  Planta.  51:  518-543. 

Pisek,  A.;  Winkler,  E.  1959.  Licht-  und  Temperaturabhan- 


gigkeit  der  COg-Assimilation  von  Fichte  (Picea  excelsa 
Lind. ),  Zirbe  (Pinus  cembran  L. )  und  Sonnenblume 
(Helianthus  annuus  L.  ).  Planta.  53:  532-550. 

Pisek,  A.;  Larcher,  W.;  Moser,  W.;  Pack,  I.  1969. 
Kardinale  Temperaturbereiche  der  Photosynthese 
und  Grenztemperaturen  des  Lebens  der  Blatter  ver- 
schiedener  Spermatophyten.  III.  Temperatur- 
abhangigkeit  und  optimaler  Temperaturbereich  der 
Netto-Photosynthese.  Flora  B.  158(3):  608-630. 

Pisek,  A.;  Larcher,  W.;  Unterholzner,  R.  1967.  Kardinale 
Temperaturbereiche  der  Photosynthese  und  Grenz- 
temperaturen des  Lebens  der  Blatter  verschiedener 
Spermatophyten.  I.  Temperaturminimum  der  Netto- 
assimilation,  Gefrier-  und  Frostschadenbereiche  der 
Blatter.  Flora  B.  157(1):  239-264. 

Pisek,  A.;  Larcher,  W.;  Unterholzner,  R.;  Pack,  I.  1968. 
Kardinale  Temperaturbereiche  der  Photosynthese  und 
Grenztemperaturen  des  Lebens  der  Blatter  verschie- 
dener Spermatophyten.  II.  Temperaturmaximima  der 
Netto-Photosynthese  und  Hitzeresistenz  der  Blatter. 
Flora  B.  158(2):  110-128. 

Rikli,  M.  1909.  Die  Arve  in  der  Schweiz.  Neue  Denkschr. 
Schweizerische  naturforschende  Gesellschaft.  44.  455  p. 

Schonenberger,  W.;  Frey,  W.,  eds.  1988.  Untersuchungen 
zur  Okologie  imd  Technik  der  Hochlagenauffbrstimg. 
Forschimgsergebnisse  aus  dem  Lawinenanrissgebiet 
Stillberg.  Schweizerische  Zeitschrift  fiir  Forstwesen. 
139(9):  735-820. 

Schwarz,  W.  1970a.  Der  Einfluss  der  Photoperiode  auf 
das  Austreiben,  die  Frostharte  und  die  Hitzeresistenz 
von  Zirben  und  Alpenrosen.  Flora.  159:  258-285. 

Schwarz,  W.  R.  1970b.  Frost  hardiness,  resistance  to  heat, 
and  COg  uptake  in  seedlings  and  mature  stages  of 
Pinus  cembran  L.  Proc.  3.  Forest  Microclimate  Sympo- 
sium, Kananaskis  Forest  Experimental  Station;  1969; 
Seebe,  AB:  137-145. 

Schwarz,  W.  1971.  The  photos)nithetic  capacity  of  some 
evergreens  during  the  winter  and  the  speed  of  their 
recovery  of  activity  after  sharp  frosts.  Berichte  der 
deutschen  botanischen  Gesellschaft.  84(10):  585-594. 

Simony,  F.  1870.  Die  Zirbe.  Jahrbuch  des  osterreichischen 
Alpenvereins.  6:  349-359. 

Tranquillini,  W.  1955.  Die  Bedeutimg  des  Lichtes  und  der 
Temperatur  fiir  die  Kohlensaureassimilation  von  Pinus 
cembran-tJungwuchs  an  einem  hochalpinen  Standort. 
Planta.  46:  154-178. 

Tranquillini,  W.  1957.  Standortsklima,  Wasserbilanz  und 
COg-  Gaswechsel  junger  Zirben  (Pinus  cembran)  an  der 
alpinen  Waldgrenze.  Planta.  49:  612-661. 

Tranquillini,  W.  1958.  Die  Frostharte  der  Zirbe  imter 
besonderer  Beriicksichtigung  autochoner  imd  aus 
Forstgarten  stammender  Jimgpflanzen. 
Forstwissenschaftliches  Centralblatt.  77:  89-105. 

Tranquillini,  W.  1959a.  Die  Stoffproduktion  der  Zirbe  an 
der  Waldgrenze  wahrend  eines  Jahres.  Planta.  54: 
107-129. 

TranquiUini,  W.  1959b.  Die  Stoffproduktion  der  Zirbe 
(Pinus  cembran  L. )  an  der  Waldgrenze  wahrend  eines 
Jahres.  II.  Zuwachs  imd  COg-Bilanz.  Planta.  54(2): 
130-151. 


65 


Tranqmllini,  W.  1963a.  Der  Jahresgang  der  CO^- 
Assimilation  junger  Zirben.  Mitteilungen  der  forst- 
lichen  Bundesversuchsanstalt,  Wien.  60:  501-534. 

Tranquillini,  W.  1963b.  Die  COg-Jahresbilanz  und  die 
Stoifproduktion  der  Zirbe.  Mitteilungen  der  forstlichen 
Biindesversuchsanstalt,  Wien.  60:  535-546. 

Tranquillini,  W.  1963c.  Die  Abhangigkeit  der  Kohlen- 
saureassimilation  junger  Larchen,  Fichten  und  Zirben 
von  der  Luft-  und  Bodenfeuchte.  Planta.  60:  70-94. 

Tranquillini,  W.  1963d.  Uber  die  Frostresistenz  der  Zirbe. 
Mitteilungen  der  forstlichen  Bundesversuchsanstalt, 
Wien.  60:  547-562. 

Tranquillini,  W.  1965.  Uber  den  Zusammenhang  zwischen 
Entwicklungszustand  und  Diirreresistenz  junger  Zirben 
{Pinus  cembran  L.)  im  Pflanzgarten.  Mitteilungen  der 
forstlichen  Bundesversuchsanstalt,  Wien.  66:  241-271. 

Tranquillini,  W.  1979.  Physiological  ecology  of  the  alpine 
timberline.  Berlin:  Springer- Verlag.  137  p. 


Tranquillini,  W.;  Machl-Ebner,  I.  1971.  Uber  den  Einfluss 
von  Warme  auf  das  Photosynthesevermogen  der  Zirbe 
(Pinus  cembran  L.)  und  der  Alpenrose  (Rhododendron 
ferrugineum  L.)  im  Winter.  Report  of  the  Kevo  Subarc- 
tic Research  Station.  8:  158-166. 

TranquilUni,  W.;  Schiitz,  W.  1970.  Uber  die  Rinden- 
atmimg  einiger  Baume  an  der  Waldgrenze.  Central- 
blatt  fiir  das  gesamte  Forstwesen.  87:  42-60. 

Turner,  H.;  Streule,  A.  1983.  Wurzelwachstum  iind 
Sprossentwicklung  junger  Koniferen  im  Klimastress 
der  alpinen  Waldgrenze,  mit  Berucksichtigung  von 
Mikroklima,  Photosynthese  und  Stoffproduktion. 
Irdning,  Austria:  Bundesanstalt  Gimipenstein:  617-635. 

Ulmer,  W.  1937.  Uber  den  Jahresgang  der  Frostharte 
einiger  immergriiner  Arten  der  alpinen  Stufe,  sowie  der 
Zirbe  und  Fichte.  Jahrbiicher  fur  wissenschaftliche 
Botanik.  84:  553-592. 


66 


A  PATTERN  OF  PINUS  PUMILA  SEED 
PRODUCTION  ECOLOGY  IN  THE 
MOUNTAINS  OF  CENTRAL 
KAMTCHATKA 

Peter  A,  Khomentovsky 


Abstract— This  paper  discusses  the  results  of  Pinus  pumila  cone 
production  studies  in  the  mountains  of  central  Kamtchatka.  Pi- 
nus pumila  was  found  to  have  a  very  high  seed  production  poten- 
tial when  not  overtopped  by  neighboring  trees.  When  shaded, 
few  or  no  cones  were  produced  on  P.  pumila.  Seed  production 
was  not  significantly  dependent  on  altitude  and  site  conditions. 
However,  elevations  above  1,300-1,400  m  were  unfavorable  for 
cone  survival,  mainly  because  of  low  temperature  and  phenologi- 
cal  delay.  Successful  seed  production  requires  snow  cover  to 
protect  ovulate  buds  on  low-lying  branches  from  freezing  in  the 
winter.  In  the  summer,  good  seed  production  is  related  to  high 
insolation,  wind  protection  on  cold  sites,  absence  of  shading,  and 
good  drainage  for  the  root  systems.  At  midelevations,  under  good 
seed  production  conditions  in  Kamtchatka,  about  110  kg  of  seeds 
per  hectare  will  be  produced  as  compared  to  only  1  kg  at  the  up- 
per elevational  limits.  Pinus  pumila  appears  to  have  a  high  evo- 
lutionary potential  that  reveals  itself  in  prominent  and  wide 
polymorphism  of  reproductive  organs. 


The  main  purpose  of  this  paper  is  to  give  a  brief  descrip- 
tion of  some  reproductive  characteristics  and  ecological 
features  of  typical  subalpine  dwarf  pine  (Pinus  pumila 
[Pall.]  Regel)  (Pinaceae,  Strobus)  (Pp)  for  many  regions 
of  its  area  ecotopes  of  the  central  Kamtchatka  Mountains 
in  northeastern  Asia  (fig.  1). 

Some  years  ago,  we  tried  to  identify  the  response  of 
seed  production  (reflected  in  cones,  seed,  and  nuclei  masses 
and  sizes)  to  some  of  the  most  evident  differences  of  Pp 
environments,  mainly  to  ecotopes,  altitudinal  position, 
and  their  location  in  regions  with  maritime  or  continental 
climate.  We  noticed  (Khomentovsky  and  Khomentovskaya 
1990)  that  there  is  no  strong  dependence  of  seed  param- 
eters and  cone  crop  on  the  ecotope,  elevation,  and  its  posi- 
tion in  the  relief,  Pp  was  supposed  to  have  some  ecotopic 
(macro-  or  microclimatic)  optimum  of  seed  production, 
which  could  be  indirectly  characterized  by  the  climate 
within  the  belt  of  600  to  900  m  above  sea  level  in  the  inte- 
rior continental  climate  regions  of  Kamtchatka  peninsula, 
and  widened  in  its  lower  limit  almost  to  sea  level  near  the 
coast. 

At  the  same  time  we  tried  to  develop  some  of  the  first 
descriptions  of  conophagous  insects  feeding  on  Pp  and  to 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Peter  A.  Khomentovsky  is  Head  of  the  Biogeocenology  Laboratory, 
Kamtchatka  Institute  of  Ecology  and  Nature  Management,  Prospect 
Rybakov,  19-A,  Petropavlovsk-Kamtchatsky  683024  Russia. 


define  their  role  in  cone  crop  size  and  quality  variation. 
We  found  (Khomentovsky  and  Efremova  1991)  that  the 
only  two  presently  known  insect  species  inhabiting  Pp 
cones  are  Cecidomyia  pumila  Mamaev  and  Efi-emova 
(Diptera,  Cecidomyidae)  and  Eupithecia  abietaria  Groeze 
(Lepidoptera,  Geometridae)  (the  taxonomy  of  the  last  spe- 
cies has  to  be  rechecked),  and  they  affect  crop  quality  very 
little.  Therefore,  this  damage  could  be  treated  as  negli- 
gible or  at  least  not  proven  to  be  important. 

These  preliminary  conclusions  forced  us  to  proceed  with 
a  more  detailed  investigation  of  Pp  seed  production.  Hav- 
ing no  possibilities  for  obtaining  direct  microclimatic  in- 
formation, we  decided  to  define  possible  principal  responses 
of  Pp  seed  production  to  elevation  and  ecotopical  features 
within  a  small  creek  basin  with  a  high  altitudinal  gradient. 

THE  RESEARCH  AREA 

The  research  area  is  situated  almost  in  the  geographi- 
cal center  of  the  Kamtchatka  (56  °N.,  158  °E.),  in  the 
Sredinnij  Mountain  Range  system,  the  main  system 
in  the  peninsula  (fig.  1).  According  to  the  botanical- 
geographical  regionalization  scheme  of  L.  F.  Kimitsin 
(1963),  this  territory  is  included  in  the  "mountain  above 
alpine  tundra  shrubby  region."  According  to  oar  scheme 
(Khomentovsky  and  others  1989),  it  lies  in  the  contact 
area  of  "Central  Kamtchatka  plainland-submountain 
province  of  conifer-stonebirch  forests"  and  "Sredinno- 
Western  middle  mountain  stone  birch-tundra  forest  prov- 
ince." The  area  covers  the  upper  part  of  the  conifer  forest 
vegetation  belt,  subalpine,  alpine,  and  above-alpine  belts 
in  Kamtchatka  (figs.  1  and  2). 

Climate  is  subcontinental,  with  300  to  400  mm  of  an- 
nual precipitation,  moderately  cold  and  snowy  winters, 
and  cool  summers  (monthly  average  air  temperature  in 
Esso  village  is  3  °C  in  May,  9.5  °C  in  June,  13  °C  in  July, 
12  °C  in  August,  and  6  °C  in  September.  The  permanent 
snow  cover  at  elevations  of  700  m  and  higher  begins  to 
accumulate  in  the  first  half  of  October  and  disappears 
mEiinly  in  May. 

Orographically,  the  area  can  be  called  "typical  middle 
altitudinal-mountainous"  (according  to  common  under- 
standing), which  for  Kamtchatka  actually  means  middle 
and  high  mountains  (due  to  the  compression  of  its  altitu- 
dinal vegetation  belts).  Average  height  of  the  mountains 
is  about  1,500  m.  Geological  and  geomorphological  struc- 
ture of  the  area  is  determined  by  volcanogenic  basalts  and 
andesits  (Ql,  Q2,  Q4),  and  by  quatenary  glacial,  fluvio- 
glacial  (Q3),  alluvial,  and  proluvial  (Q4)  deposits.  Glacial 
relief  is  conspicuous,  especially  in  the  upper  parts  of  river 


67 


I 


Figure  1 — General  description  of  the  research  area  in  Kamtchatka 
peninsula.  Top  left:  principal  scheme  of  vegetation  belts:  1 — subalpine, 
2 — boreal  forest,  3 — alpine,  4 — coastal  and  lowland.  Top  right:  main 
mountain  ridges  and  hydrological  net.  Bottom  left:  research  area,  within 
small  basin.  Dots  are  points  where  material  was  collected. 


more  or  less  permanent  (on  large  time  scale)  weak  or 
moderate  volcanic  ashfalls  impacts  (mainly  from  the  east- 
ern part  of  peninsula). 

Due  to  the  geographical  position,  mountain  relief,  and 
severe  climate  of  the  area,  vegetation  is  not  very  diverse 
and  exhibits  a  mosaic-like  spatial  pattern.  River  valleys 
and  lower  parts  of  their  tributaries  are  occupied  by  Popu- 
lus  suaveolens,  Chosenia  arbutifolia,  Larix  cajanderi, 


and  creek  basins.  Soils  are  typical  for  the  Kamtchatka 
Mountain  regions  classification  (I.  A.  Sokolov  1973),  with 
the  additions  by  N.  V.  Kazakov  (personal  communication) 
They  include  turf  illuvio-humus,  turf  illuvio-volcanic  ash, 
tundra  illuvio-humus,  tundra  illuvio-humus- volcanic  ash, 
and  turf  primitive.  Soil  genetics  reveal  themselves  in  the 
existence  of  specific  features  including  the  long-term  fi'eez- 
ing  period,  anaerobic  reactions  in  the  lower  horizons,  and 


68 


Figure  2 — A  fragment  of  the  vegetation  cover  in  the  upper  part  of  the  altitudinal  transect  (see  also  fig.  1). 


Betula  kamtschatica,  andB.  ermanii.  Two  dwarf  tree  spe- 
cies, Pinus  pumila  audi  Alnus  kamtschatica,  are  widely 
spread  along  the  whole  vegetation  profile.  They  occur  by 
groups  or  strips  along  the  moraine  tongues,  river  terraces, 
fluvioglacial  deposits,  and  watershed  ridges  and  slopes, 
and  form  the  upper  limit  of  woody  vegetation  (fig.  2). 

Plant  formations  in  the  middle  and  high  altitudes  are 
almost  the  same  (except  for  Populus  and  Chosenia),  but, 
moving  upward  along  the  narrowing  valleys  and  water- 
sheds, along  the  altitudinal  environmental  gradient  of 
gradually  decreasing  plant  life  conditions,  we  can  see 
the  distribution  pattern  of  the  formation  changing  from 
"macromosaic"  to  "micromosaic."  Tree  distribution  is 
more  strictly  related  to  soil  temperature  conditions  and 
drainage  gradients,  and  to  sites  sheltered  from  wind  and 
snow  abrasion.  This  is  true  even  for  such  ecologically 
flexible  dwarf  species  as  Pinus  pumila  and  Alnus 
kamtschatica. 

The  upper  elevational  limits  of  upright-grovnng  woody 
vegetation  {Larix  kamtchatica,  Betula  ermanii)  are  lo- 
cated between  900  to  1,000  m,  depending  on  site  condi- 
tions. Pinus  pumila  and  Alnus  kamtschatica  occur  up  to 
1,300  to  1,400  m  (rare  cl\mips).  Up  to  1,200  m  P.  pumila 
is  able  to  produce  seeds. 

MATERIALS  AND  METHODS 

Data  were  gathered  mostly  in  September  to  October  1990 
(with  some  additional  data  from  other  years)  at  the  field 
base  "Bolgit,"  in  the  vicinity  of  Esso  village  (Bistrinsky 
district  of  Kamtchatka  region),  in  the  "Tupikin  clyutch" 
creek  basin,  within  an  area  of  about  10  km^.  This  paper 
presents  the  material  from  one  slope  exposed  to  the  east- 
northeast  and  from  the  area  nearest  to  the  watershed  (up 
to  50  m  from  the  slope  edge),  within  an  altitudinal  range 
of  650  to  1,030  m.  Permanent  plots  on  an  existing  tran- 
sect and  some  additional  temporary  points  were  chosen 
for  the  cone  crop  measuring  and  sampling  (fig.  1). 

The  follovmig  information  was  collected  on  each  sample 
site  (table  1):  type  of  plant  commimity,  topographic 


location  and  ground  cover  of  Pp  clumps,  average  seasonal 
height  of  skeleton  branches,  general  moisture  conditions, 
shading  by  neighboring  upright  trees  or  climips. 

EXPLANATORY  COMMENTS 

First,  strictly  speaking,  we  cannot  consider  Pp  plants 
as  common  trees — they  are  not  standing  separately  and 
do  not  have  a  single  trunk;  in  most  cases  they  overlap 
each  other,  both  aboveground  and  in  imderground  parts. 
Also,  we  cannot  call  them  "clones"  because  we  do  not  know 
their  origin  and  level  of  genetic  relationships — which  is 
especially  important  in  the  case  of  seed  dispersal  by  ani- 
mals. The  best  term  would  be  "clump"  (we  are  grateful  to 
Dr.  Diana  Tomback  for  some  explanations  on  this  topic). 

Second,  the  single  trunk  (in  traditional  understanding) 
of  Pp  is  considered  a  creeping  tree  not  a  bush,  and  its 
length  is  only  several  centimeters  or  tens  of  centimeters. 
Branches  of  first  and  second  order  (I  prefer  to  call  them 
"trunk-branches")  exhibit  the  same  physiognomy,  bear 
similar  quantity  of  shoots,  and  form  multi trunk-like 
crovms.  In  the  case  of  syngenetic  origin  and  continuous 
dispersal,  Pp  plants  form  such  dense  cover  by  overlapping 
basal  parts  in  the  litter  and  root  systems  that  it  is  impos- 
sible to  identify  single  plants  and  thus  their  total  num- 
bers without  destroying  the  stand.  This  forced  us  to  use 
the  only  acceptable  method  of  measuring  cone  crops  or 
similar  parameters — instead  of  measuring  only  one  tree 
or  clump,  measure  by  square  unit  such  as  hectares  or 
square  meters.  Certainly,  the  stand  structure  and  charac- 
teristics of  groimd  cover  have  to  be  taken  into  account, 
especially  for  practical  taxation. 

Third,  it  is  also  impossible,  in  many  cases,  to  determine 
real  age  of  adult  Pp  plants.  They  produce  adventitious 
roots,  are  growing  apically,  and  are  gradually  decaying 
in  their  basal  parts  during  most  of  their  lifetime,  theoreti- 
cally endlessly.  We  have  to  recognize  that  we  really  can 
only  know  the  age  at  the  moment  of  measuring,  similar 
to  being  counted  in  young  trees,  and  not  the  real  one, 
which  is  usually  higher. 


69 


Table  1 — Short  description  of  the  ecotopes  and  Pinuspumila  clumps  being  analyzed.  The  age  of  clumps  in  these  analyses  lies  between  150 
and  260  years' 


Altitude, 

Pinus  pumila 

Way  of 
main  water 

Woodstand 

Pinus  pumila 

Shading 

m  asl 

Position  In  the  relief 

community  type 

supply 

composition 

H  avg,  cm 

SO,  % 

UTr 

Ppc 

1,030 

Upper  part  of  the  flat 
watershed,  shaded  from 
S  and  E  by  the  ridge 

Pumilae  pinetum 
carioso-hypnoso- 
ericosum,  with 
fragments  of  P.p. 
cladinosum  (CHE*C^ 

A+S 

lOPp 

100  (15-20) 

80  unev 

0 

1 

950 

Plateau,  opened  to  all 
directions,  watershed 
with  slight  slope  to 
NNE 

P.p.  purum,  with 
fragments  of 
P.p.  carioso- 
cladinosum  (P*CC^ 

A 

lOPp 

40  (10-15) 

40  unev 

0 

0 

900 

Flat  watershed  above 
the  creek  source, 
shaded  by  the  ridge 
from  S 

P.p.  hypnoso- 

carioso-ericosum 

(HCE) 

A+S 

lOPp+Lk 
(upper 
limit  of 
Lk) 

300 

40  unev 

1 

1 

810 

A  ridge  of  the  complex 
watershed,  slight 
slope  to  the  NE 

P.p.  carioso- 

hypnoso-ericosum 

(CHE) 

A 

lOPp+Lk 

150  (35-40) 

60  unev 

1 

2 

800 

Middle  part  of  the  E 
exposed  slope  of  wide 
creek  valley 

P.p.  hypnoso- 

carioso-ericosum 

(HE) 

S+A 

lOPp+Lk 

200 

80 

0 

1 

680 

Eastern  border  of  the 
watershed  ridge  with 
flattened  top 

P.p.  hypnoso- 

carioso-ericosum 

(HCE) 

A+S 

7Pp  3Lk 

300 

60  unev 

2 

1 

650 

Lower  part  of  E 
exposed  slope  in 
narrow  creek  valley 

P.p.  ericoso- 
sphagnosum 

8 

lOPp+Ak 

300  (40-45) 

100 

0 

3 

^ Pinus  pumila  community  type:  in  parentheses — abbreviation  for  fig.  8;  Way  of  main  water  supply:  A — atmospheric,  S — slope;  Woodstand  composition:  Pp — Pinus 
pumila,  Lk — Larix  cajanderi,  Ak — Alnus  kamtschatica  (dwarf  alder);  Pinus  pumila:  H  avg — average  height  (cm),  in  parentheses — rough  estimation  of  annual  shoot 
elongation  (mm);  SC% — surface  covering  by  the  clumps,  "unev" — uneven;  Shading:  Utr— from  upright  trees  of  different  species,  Ppc — from  Pinuspumila  neighbor- 
ing clumps;  0 — no  shading,  1 — slight,  2 — moderate,  3 — heavy  shading. 


In  each  point  for  data  and  sample  collecting  v^^e  used 
3  to  5  small  randomized  plots  of  different  size — from  1  to 
15  m^.  The  following  parameters  were  countered  in  place 
(table  2):  number  of  skeleton  branches  ("trunk-branches"), 
number  of  germinating  shoots  on  each  skeleton  branch; 
number  of  current  year  crop  cones  (the  second  year  of 
cone  development)  and  number  of  next  year  crop  cones 
(female  buds  of  the  first  year)  on  each  germinating  shoot 
and  each  skeleton  branch;  and  among  the  total  quantity  of 
each  year's  cones  the  number  of  cones  damaged  by  insects 
iCecidomyia  pumila — damage  in  the  first  year  of  cone  de- 
velopment, Eupithecia  abietaria — damage  in  the  second 
year  of  cone  development),  and  by  the  nutcracker  {Nuci- 
fraga  caryocatactes  kamtschatkensis) . 

At  each  site,  10  to  100  cones  of  the  current  year's  crop 
were  collected  and  measured  before  they  dried  (cone 
length  and  cone  diameter).  Later,  after  air  drying  during 
some  weeks  or  months,  cones  and  seeds  were  measured 
by  16  other  parameters,  six  of  which  are  discussed  in  this 
paper:  cone  mass;  1,000  seeds  mass;  1,000  nuclei  mass 
(both  by  measuring  a  number  of  samples  with  50  seeds  in 
each);  nimiber  of  seed  scales  in  the  cone;  among  the  total 


number  of  seed  scales,  the  number  of  scales  not  contain- 
ing seeds  under  them;  and  total  seed  quantity  in  the  cone. 

RESULTS  AND  BRIEF  COMMENTS 

•  The  structure  of  the  cone-bearing  portion  of  the 
crown  and  cone  crop  estimates  (tables  1  and  2,  figs.  1-6): 

As  a  whole,  the  number  of  skeleton  branches  and  germi- 
nating shoots  per  hectare  varied  very  little  within  alti- 
tudes of  650-800  m  and  definitely  increased  upward.  This 
could  be  explained  by  increasing  isolation.  Most  likely  it 
was  the  reason  for  the  especially  increasing  number  of 
germinating  shoots  (two  to  three  times  higher  than  skel- 
eton branches)  and,  correspondingly,  the  nvimbers  of 
current-year  and  next-year  cones.  The  last  varied  more 
than  the  first  because  some  of  them  vdll  inevitably  die  in 
the  natural  selection  process  during  the  maturation  time 
in  the  coming  year. 

The  number  of  shoots  with  cones  per  skeleton  branch 
(fig.  4)  changed  opposite  to  that  of  the  number  of  skeleton 
branches  per  1  hectare  (we  noticed  the  increased  number 


70 


of  shoots,  including  nongerminating).  Even  a  slight  plant 
shading  immediately  resulted  in  a  change  in  the  number 
of  shoots  (sites  900  m  and  1,030  m). 

The  quantity  of  current  and  next-year  crop  cones  per 
one  germinating  shoot  had  specific  dispersal  features: 
at  each  site,  especially  in  the  lower,  shaded  part  of  the 
elevation  profile,  cones  of  any  one  year  were  overwhelm- 
ing. This  migration  of  intensive  seed-producing  centers 
from  year  to  year  supports  our  previous  conclusion 


(Khomentovsky  and  Efremova  1991)  about  the  existence 
of  local  seed  production  mosaics,  which  provides  a  con- 
tinuous supply  of  seeds,  important  not  only  for  guaran- 
teed reproduction  and  microevolutionary  diversity  for  Pp, 
but  also  for  all  zooconsumers  such  as  nutcracker,  other 
birds,  and  mammals. 

•  Cones  and  seeds  mass  and  size  variation  (tables  1,  3; 
figs.  7  and  8): 


Table  2 — Characteristics  of  germinating  crown  parts  and  cone  crop  outcome  of  Pinus  pumila  in  central  Kamtcliatka  Mountains^ 


nn 

Parameter 

Altitude  (m  above  sea  level) 

650 

680 

800 

810 

900 

950 

1,030 

1 

No.  of  skeleton  branches/ha 

1,200 

4,167 

2,240 

21,111 

11,000 

30,000 

10,333 

2 

No.  of  germinating  shoots/ha 

1,733 

7,917 

3,200 

47,778 

57,000 

86,667 

33,000 

3 

No.  of  this  year  crop  cones/ha 

1,733 

4,375 

3,136 

2,389 

46,170 

56,334 

28,000 

4 

No.  of  next  year  crop  cones/ha 

260 

0 

160 

51,122 

19,950 

143,000 

24,550 

5 

No  of  aerminatina  shoots/skel  bran 

1.44 

1.90 

1.43 

2.26 

5.19 

2.89 

3.00 

6 

No  of  thi^  vppir  oroo  conp^/<>kpl  hri^n 

1.44 

2.00 

1.40 

0.11 

4.20 

1.88 

2.56 

7 

No  of  next  vear  croo  cones  /skel  bran 

0.22 

0 

A  A7 

O  AO 

1  Md. 

4.77 

2.54 

o 
O 

No.  of  this  year  crop  cones/germ,  shoot 

1 .00 

H  AC 
1  .00 

0.98 

0.05 

0.81 

A 

A  7ft 

9 

No.  of  next  year  crop  cones/germ,  shoot 

0.15 

0 

0.05 

1.07 

0.35 

1.65 

0.86 

10 

Percentage  of  skeleton  branches  with 

various  numbers  of  germinating  shoots: 

with  one  shoot 

67 

30 

64 

58 

28 

22 

32 

with  two  shoots 

22 

CA 

50 

29 

16 

18 

J4 

with  thrfifi  shoots 

will  1   11  II           Ol  Iwwlw 

11 

20 

7 

16 

0 

22 

8 

with  four  shoots 

0 

0 

0 

0 

9 

11 

4 

with  fiv/o  chnr^tQ 
Will  i  live?  Ol  IwUlO 

0 

0 

0 

5 

9 

0 

8 

with  qIy  QhnntQ 

Willi  OlA  Ol  lUUlO 

0 

0 

A 
0 

A 
0 

A 
0 

0 

2 

with  seven  shoots 

0 

0 

0 

0 

0 

0 

5 

with  eight  shoots 

0 

0 

0 

0 

9 

11 

5 

with  nine  shoots 

0 

0 

0 

0 

g 

0 

0 

with  ten  shoots 

0 

0 

0 

0 

0 

0 

2 

with  eleven  shoots 

0 

0 

0 

0 

0 

0 

0 

with  twelve  shoots 

0 

0 

0 

5 

0 

0 

0 

with  thirteen  shoots 

0 

0 

0 

0 

Q 

0 

1 

1 1 

r  ciccf iictgc  ui  gcniiiriaurig  biiouis  Dcaiirig 

various  number  of  this  year  cones  (T) 

and  next  year  cones  (N): 

0N-1T 

81 

100 

93 

86 

63 

8 

32 

ON -21 

7 

0 

2 

8 

4 

0 

6 

Cumulative  percent  of  shoots  with 

this  year's  crop  cones  only 

88 

100 

95 

94 

67 

8 

38 

1N-0T 

4 

0 

0 

3 

25 

15 

30 

2N-0T 

4 

0 

5 

0 

2 

11 

7 

3N-0T 

0 

0 

0 

0 

0 

11 

0 

Cumulative  percent  of  shoots  with 

this  year's  crop  cones  only 

8 

0 

5 

3 

27 

37 

37 

1N-1T 

4 

0 

0 

3 

3 

23 

10 

1N  -2T 

0 

0 

0 

0 

3 

4 

2 

2N  -  1T 

0 

0 

0 

0 

0 

15 

6 

2N-2T 

0 

0 

0 

0 

0 

5 

5 

3N-1T 

0 

0 

0 

0 

0 

8 

1 

4N  -  1T 

0 

0 

0 

0 

0 

0 

1 

Cumulative  percent  of  shoots  with 

cones  of  both  years 

4 

0 

0 

3 

6 

55 

25 

'Only  the  area  covered  with  P'mus  pumila  clumps  is  taken  into  account  (par.  1  -4),  not  the  whole  territory  of  the  site;  number  of  next-year  cones  can  decrease 
during  the  year  due  to  natural  selection  processes. 


71 


Table  3 — Some  characteristics  of  Pinus  pumila  cones  and  seeds  in  various  elevations  in  central 
Kamtchatka  Mountains  (within  research  area)' 


Regression  on  altitude  Correlation  Average  value 


CL 

=  54.71 



0.020  A 

R 

=  -0.73 

AVG  (cl)  = 

43 

+/- 

8  (25-62) 

CD 

=  33.68 

- 

0.010  A 

R 

=  -0.79 

AVG  (cd)  = 

27 

+/- 

2(18-37) 

CM 

=  7.81 

0.030  A 

R 

=  -0.41 

AVG  (cm)  = 

7 

+/- 

1  (4-10) 

SSN 

=  32.30 

+ 

0.010  A 

R 

=  0.53 

AVG(ssn)  = 

39 

+/- 

5  (26-52) 

SWS 

=  9.06 

+ 

0.004  A 

R 

=  0.33 

AVG  (SWS)  = 

11 

+/- 

3  (5-18) 

SQT 

=  28.50 

+ 

0.003  A 

R 

=  0.76 

AVG  (sqt)  = 

45 

+/- 

5  (26-68) 

SMP 

=  46.71 

+ 

0.002  A 

R 

=  -0.11 

AVG(smp)  = 

45 

+/- 

3  (32-55) 

SMTH 

=  101.67 

0.020  A 

R 

=  -0.45 

AVG  (smth)  = 

84 

+/- 

8  (52-116) 

NMTH 

=  56.49 

0.020  A 

R 

=  -0.49 

AVG  (nmth)  = 

43 

+/- 

4  (27-68) 

'CL — cone  length,  mm;  CD — cone  diameter,  mm;  CM — cone  mass,  g;  A — altitude,  m  above  sea  level;  SSN — seed 
scale  number;  SWS— number  of  scales  without  seeds  (within  SSN);  SQT — total  number  of  seeds  in  the  cone;  SMP — 
percentage  of  seeds  mass  in  cone  mass;  SMTH — mass  of  1 ,000  seeds,  g;  NMTH — mass  of  1 ,000  nuclei,  g.  In  paren- 
theses at  the  right  side — minimum  and  maximum  values. 


Mass  variation  (here  and  later  we  speak  about  current- 
year  crop  cones  only)  is  more  evident  in  the  upper  part  of 
the  elevation  profile,  in  micromosaic  sites  with  increasing 
abiotic  environmental  pressure.  In  sufficiently  insolated 
(sunny)  and  wind-protected  sites  cone  mass  is  increasing, 
in  shaded  sites  or  windy  plateaus — decreasing.  At  the  up- 
per elevation  levels  of  Pp  distribution  (1,200-1,300  m)  we 
observed  natural  abrupt  decreases  of  all  seed-producing 
parameters.  It  is  remarkable  that  in  general  cones  and 
seeds  masses  are  not  too  closely  related  to  definite  altitu- 
dinal  or  site  positions  of  the  tree,  even  in  cases  of  evi- 
dently unfavorable  envirormients  (site  950  m). 

Self  regulation  of  seed  production  is  obvious  also:  along 
the  elevation  gradient,  number  of  seed  scales  in  the  cone 


(with  and  without  seeds)  and  total  number  of  seeds  in  the 
cone  slightly  increases  as  well,  and  at  the  same  time  seed 
and  nuclei  mass  is  unchanging  or  slightly  decreasing.  It 
means  that  seeds  become  smaller  but  grow  in  quantity, 
keeping  the  same  reproduction  potential  and  having  more 
protection  from  environmental  impacts. 

Cone  size  variation,  most  evident  in  their  length,  corre- 
sponds vrith  mass  variation  in  relation  to  site  position  in 
the  landscape.  Here,  as  well  as  in  the  previous  case,  the 
main  impact  factor  is  not  elevation  but  landscape  struc- 
ture of  the  ecotope  (slopes  in  wide  or  narrow  valley,  pla- 
teau, watershed  ridge,  etc.).  Biocenotic  structure  is  not 
too  important  in  this  case,  at  least  at  the  forest  type  level. 
All  ecotopes,  excluding  the  950-m  site  (opened  plateau), 


160 


140 


§  120 
O 


£  100 

(0 

u 

X  80 

CL 

O  60 

n 
E 

3 

Z  40 


20 


i 

\ 

\ 

i 
i 

\ 

\ 

— \  

— i-/- 

\ 

.  \ 

/ 
/I 

\  \ 
\\ 

 V 

\  1 

•  y 

>' 
/ 

f 

\l 
1  • 

1  . 

1  / 

1  ^ 

 skeleton  branches 

 germinating  shoots 

—  current  year  cones 
 next  year  cones 


650    680    800    810    900    950  1.030 

Altitude  (m  as!) 

Figure  3 — Principal  structure  of  P'mus  pumila  crowns  at 
various  altitudes  and  estimated  cone  crop  per  hectare. 


72 


6 


germ,  shoots/sk.  br. 


650        680        800         810        900         950  1,030 


Altitude  (m  asl) 

Figure  4 — Structure  of  Pinus  pumila  crown  germinating 
part  (number  of  germinating  shoots  and  estimated  cone 
crop  per  skeleton  branch). 


are  occupied  by  the  same  group  of  forest  types — Pumilae 
pinetum  carioso-hypnoso-ericosum,  and  its  variations. 
Generally  speaking,  cones  size,  as  well  as  cones  mass, 
which  has  good  correlation  with  them  (r  =  0.7-0.8),  vary 
within  limits  already  known  for  the  species. 

•  Damage  caused  by  insects  and  birds  (fig.  9): 

This  was  investigated  in  some  detail  previously 
(Khomentovsky  and  Efremova  1991)  and  is  not  likely 


to  effect  any  substantial  change  in  average  cone  sizes:  it 
does  not  affect  cone  diameter  and  leads  to  cone  length  de- 
creases up  to  3-9  percent  (Cecidomyia  pumila).  However, 
in  some  cases  this  insect  species  damage  results  in  under- 
development of  up  to  20  percent  of  the  seeds,  but  this  im- 
pact usually  does  not  stop  the  development  of  the  rest  of 
the  seeds  in  the  cone  and  is  compensated  at  the  popula- 
tion level  by  the  abundant  production  of  seeds. 


73 


100 


cones  of  both  years 
next  year  cones  only 
current  year  cones  only 


650    680  800    810    900  950  1,030 

Altitude  (m  asl) 

Figure  6 — Distribution  of  current-year  and  next-year  cone  crop  on 
germinating  shoots. 


The  level  of  damage  caused  by  Cecidomyia  pumila, 
clearly  noticeable  visually  by  the  cone  curving,  is  usually 
rather  high  everywhere  but  it  varies  with  different  site, 
environmental,  and  weather  conditions,  to  which  insects 
are  more  sensitive.  In  our  case  it  was  changing  from  mod- 
erate damage  of  25-50  percent  on  the  650-m  site  (slope 
in  the  narrow  valley),  the  800-m  site  (slope  in  the  wide 


valley),  and  the  900-m  site  (wind-protected  watershed  at 
the  upper  limit  of  Larix)  to  a  maximum  of  100  percent  on 
the  810-m  site  (watershed  ridge)  and  a  minimal  8  percent 
on  the  1,030-m  site  (weakly  wind  protected  and  too  cold 
for  insects  ecotope). 

Confirmation  of  the  negligible  character  and  amount  of 
insect  damage  is  also  seen  in  the  level  of  seed  extraction 


o 


o 
m 

w 

"O 

<u 
a> 
0) 
o 
in 

V) 

<a 
c 
o 
O 


650 


800 


900 


1,030 


O 

o 

CM 
II 

o 

(/) 


Altitude  (m  asl) 


cone  mass 
SM  std 


  CM  std 

  50  nuclei  mass 


50  seeds  mass 
NMstd 


Figure  7 — Variations  in  Pinus  purpila  cones,  seeds,  and  nuclei  mass. 
SL — slopes,  P — Plateau. 


74 


650  680  800  810  900  950  1,030 


Altitude  (m  asl) 

Figure  8 — Variations  in  Pinus  pumila  cone  size.  1 — forest  type  (see  table  1). 


by  the  nutcracker,  which  collects  only  full,  matured  seeds 
(fig.  9). 

DISCUSSION  AND  CONCLUSIONS 

Data  analysis  results,  combined  with  some  previous 
material,  allow  us  to  make  some  conclusions  about  Pinus 
pumila  seed  production  in  Kamtchatka,  keeping  in  mind 
the  fragmentary  nature  of  our  study,  even  though  it  was 
done  in  typical  environmental  conditions. 

1.  The  main  conclusion  is  that  P.  pumila  seed  produc- 
tion has  such  a  high  potential  in  time  and  space  that  we 
can  speak  about  the  environmental  (abiotic  and  natural 
biotic)  impact  on  it  in  the  frame  of  modification  only  but 
not  in  the  frame  of  regulation.  Environmental  stresses 
usually  cannot  cause  a  crop  crisis — trees  have  compensat- 
ing mechanisms  both  at  the  organismic  and  population 


level,  and  one  of  them  is  the  microhabitat  mosaics  of  an- 
nual seed  production. 

2.  Pinus  pumila  seed  production  is  not  (in  Kamtchatka, 
at  least)  significantly  dependent  on  altitude  and  site  con- 
ditions. Landscape  conditions  (above  facial  level)  are  the 
most  important.  However,  this  is  true  only  when  trees 
are  not  overshadowed  by  neighboring  upright  trees;  that 
is,  when  Pp  forms  an  independent  vegetational  belt  in 
more  or  less  solid  segments — in  plain  or  mountain  tundra, 
upper  forest-tundra  ecotone,  on  the  seashore  dunes,  in 
wide  river  valleys,  etc.  Pp  often  grows  as  the  lower  canopy 
of  more  or  less  dense  Larix  forests  (more  than  30  percent 
of  Larix  cover),  which  have  the  same  age  and  the  same 
syngenetic  origin  (very  often  in  pyrogenetic  ecotopes). 
These  shaded  dwarf  pine  usually  do  not  produce  seeds 
and  develop  vegetatively  only. 

3.  The  utmost  upper  limits  of  distribution  (1,300-1,400  m 
above  sea  level)  are  unfavorable  for  Pp  seed  production. 


650 


680 


800  810  900 

Altitude  (m  asl) 


950 


1,030 


Ea+Nc 


□ 


Cp 


Ea+Cp+Nc 


Nc 


Figure  9 — Percentage  of  cones  damaged  by  insects  and  birds.  Cp — Cecidomyia 
pumila;  Ea — Eupithecia  abietaria;  Nc — Nucifraga  caryocatactes. 


75 


mainly  because  of  low  temperature  and  phenological  de- 
lay. But  even  here,  some  cones  can  be  found,  and  some- 
times the  crop  can  be  abundant  in  wind-protected  habi- 
tats. Phenological  delay  often  prevents  the  nutcracker 
from  collecting  seeds  here.  These  seeds,  having  been  dis- 
persed by  birds  from  lower  sites,  and  having  more  stable 
genotypes,  provide  a  better  chance  for  successful  refores- 
tation in  new  areas.  At  the  same  time,  dwarf  pine  regen- 
eration at  high  altitudes  is  more  or  less  provided  by  the 
remaining  abundant  seeds.  Besides  this,  numerous  zoo- 
consorts  (mainly  micromammals)  will  have  a  good  food 
supply  most  valuable  in  extreme  conditions. 

The  same  unfavorable  conditions  for  seed  production 
can  be  found  in  the  Pp  belt  at  the  seashore  (Pacific 
Kamtchatka  coast),  on  the  dunes  nearest  to  the  water. 
But  here,  as  well  as  in  the  mountains,  trees  utilize  each 
opporttmity  for  maximal  realization  of  their  reproduction 
ability.  Pp  seed  production  on  the  second  to  third  dune 
of  the  coast  belt  is  equal  in  cone  size  and  mass  parameters 
to  those  at  the  900-  to  1,000-m  altitude  and  in  seed  and 
nuclei  parameters  to  those  at  the  650-  to  750-m  altitude 
in  the  continental  climate  portion  of  the  peninsula. 

Conditions  moderately  favorable  for  Pp  seed  production 
can  be  found  in  plain  and  submoimtain  (foothill)  places, 
for  instance  on  gravel-sandy  fluvioglacial  or  proluvial 
deposits,  where  abundant  cone  crops  sporadically  occur. 

The  most  productive  seed  zone  lies  in  the  vegetation 
belt  within  the  600-900  m  above  sea  level  area.  Here, 
cone  crops  are  mostly  permanent  and  stable  from  year 
to  year. 

4.  Successful  seed  production  of  Pp  needs  the  following 
conditions:  in  winter — sufficient  snow  cover  to  protect  buds 
of  low-ljdng  trees  from  freezing;  in  summer — high  insola- 
tion, good  wind  protection  on  cold  sites,  absence  of  shad- 
ing, and  good  drainage  for  the  root  systems.  The  main 
adaptive  feature  of  Pp  is  its  ability  to  lay  close  to  the 
ground  in  the  autumn.  This  process  is  initiated  by  low 
temperature  in  the  autumn  before  the  snow  cover.  This 
unique  feature  provides  good  chances  of  survival  under 
very  extreme  conditions.  Strict  dependence  of  Pp  on  snow 
cover  depth  is  proved  by  coincidence  of  its  area  border  with 
the  snow  depth  isoline  of  40  cm  and  more  (Lukitcheva 
1964;  Tstcherbakova  1964). 

5.  Special  research  and  preliminary  work,  including 
mapping  of  various  productivity  sites,  is  needed  for  the 
practical  (silvicultural)  estimation  of  current  or  future 
cone  crops  of  Pp.  It  is  possible  to  show,  as  an  example, 
the  scope  of  crop  diversity.  In  the  middle  elevation  site 
(700-800  m),  with  5  g  of  seed  mass  in  the  cone,  80  percent 
of  germinating  shoots,  and  80  percent  of  Pp  covering  the 
ground,  we  can  get  about  110  kg  of  seeds  per  hectare.  At 
the  upper  limit  of  seed  production,  with  2  g  of  seed  mass 
in  the  cone,  30  percent  of  germinating  shoots,  10  percent 
of  covering,  we  can  get  about  1  kg  of  seeds  per  hectare. 
High  variability  of  cone  crop  is  illustrated  in  table  1  data: 
taking  into  account  only  pure  stands  of  Pp  (without  con- 
sidering its  real  distribution  on  the  surface)  we  can  see 
that  the  crop  size  on  the  650-m  site  is  about  6.5  kg/ha, 
and  about  212  kg/ha  on  the  950-m  site. 

6.  Starting  from  the  elevation  that  corresponds  to  the 
upper  limit  of  upright  trees  distribution  (900  to  1,000  m) 


in  Kamtchatka,  Pp  produces  an  increasing  number  of 
shoots.  This  can  be  interpreted  as  a  organismic  compen- 
sation for  the  increasing  severity  of  abiotic  environmental 
conditions.  Photosynthetic  ability  increases,  seed  produc- 
tion potential,  as  we  saw,  remains  at  the  same  high  level, 
and  seed  protection  (thickness  of  seed  walls)  becomes 
stronger.  All  this  takes  place  mostly  in  the  subalpine  belt 
and  proves  Pp  belongs  mainly  to  this  type  of  vegetation. 
Coincidence  of  these  activated  processes  with  the  upper 
limit  of  upright  trees  distribution  allows  us  to  hypothe- 
size the  existence  of  some  kind  of  temperature  threshold- 
trigger.  It  could  be  the  sum  of  effective  temperatures, 
above  which  the  development  of  upright  tree  forms  is  pos- 
sible, and  below  which  only  prostrate  forms  develop. 

7.  We  agree  with  the  opinion  that  Pp,  as  well  as  other 
stone  pines,  has  a  high  evolutionary  potential,  revealing 
itself  in  prominent  and  wide  polymorphism  of  reproduc- 
tive organs  and  their  functional  flexibility  (Pravdin  and 
Iroshnikov  1982).  This  is  well  illustrated  by  our  work 
and  other  sources.  Ecotypical  features  are  also  evident, 
accompanied  by  stability  of  the  main  seed-production 
parameters.  In  fact  almost  all  of  this  is  similar  to  that 
which  is  known  from  other  Pp  regions  (Bobrinev  and 
Rylkov  1984;  Efremova  and  IvHev  1972;  Kapper  1954; 
Khomentovsky  and  Khomentovskaya  1990;  Krylov  and 
others  1983;  Rush  1974). 

8.  Pp  seed  production  potential  exceeds  the  possibilities 
of  its  realization  in  our  research  area,  and  most  likely,  in 
the  Kamtchatka  peninsula  as  a  whole.  Ecotopical  diver- 
sity in  Kamtchatka  is  evidently  only  a  part  of  the  whole 
scope  of  acceptable  conditions  for  the  species  (here  we 
can  admit  that  low  temperatures  affect  vegetative  growth 
three  to  four  times  more  than  they  affect  generative  or- 
gans variability).  Therefore,  we  should  look  for  possible 
phylogenetic  roots  of  this  species  in  the  regions  with  a 
wider  geoecological  spectrum  of  sites. 

9.  Our  observations  (partly  presented  in  this  paper)  al- 
low us  to  speculate  that  Pp  is  a  middle-altitudinal  species, 
with  more  preferable  growth  conditions  in  the  moderately 
continental  climate  (in  relatively  humid  mountains)  than 
at  the  seashore.  But,  regarding  the  whole  area,  the  defi- 
nite and  necessary  influence  of  Pacific  wet  air  masses  re- 
sults in  considerable  snow  cover. 

Combining  this  conclusion  with  some  known  hypothe- 
ses of  Pp  evolution  (Sotchava  and  Lukitcheva  1953; 
Tikhomirov  1949;  etc.)  we  think  that  this  dwarf  tree  spe- 
cies most  likely  appeared  in  Tertiary  times  in  some  cli- 
matically subcontinental  mountain  regions  of  Angarida 
(northeastern  Asia).  In  Kamtchatka,  pollen  data  show 
Pp  has  been  here  for  1-1.5  million  years  (Malaeva  1967; 
Tchelebaeva  and  others  1974).  During  Pleistocene  inter- 
glacial  periods,  and  in  the  early  middle  Holocene,  Pp 
occupied  periglacial  zones  and  other  places  unacceptable 
for  more-thermophil  tree  species.  The  highest  incidence 
of  polyembriony  and  underdevelopment  of  seeds  in 
Kamtchatka  support  the  possible  centers  of  origin  in  the 
moderate  continental  climates  when  compared  to  Magadan 
(northeastern  continental  part  of  Asia)  and  Buryatia 
(southern  part  of  eastern  Siberia)  (Iroshnikov  1972;  Rush 
1974).  These  features  can  be  considered  a  reaction  to  less 
favorable  conditions  of  the  secondary  environments. 


76 


including  Kamtchatka  peninsula.  According  to  A.  I. 
Iroshnikov  (1972),  Pp  is  one  of  the  most  well-adapted 
species  to  the  continental  climate  of  northeastern  Asia. 

REFERENCES 

Bobrinev,  V.  P.;  Rylkov,  V.  F.  1984.  Seed  production  of 
dwarf  stone  pine.  Inform,  bull,  of  Tchitinsky  Informa- 
tion Center,  #76-84.  Tchita.  5  p.  [In  Russian]. 

Efremova,  L.  S.;  Ivliev,  L.  A.  1972.  On  dwarf  stone  pine 
seed  producing  at  Kamtchatka.  In:  Utilization  and  re- 
producing of  forest  resources  in  the  Far  East.  Abstr.  of 
the  all-imion  conference  Khabarovsk.  Part  2:  158-159. 
[In  Russian]. 

Iroshnikov,  A.  I.  1972.  Seed  producing  and  seed  quality 
of  conifers  in  northern  and  mountainous  regions  of 
Siberia.  In:  Seed  producing  of  Siberian  conifers  (collec- 
tion of  papers).  Novosibirsk:  98-117.  [In  Russian]. 

Kapper,  O.  A.  1954.  Conifers  (forestry  characteristics). 
IVloscow-Leningrad.  304  p.  [In  Russian]. 

Khomentovsky,  P.  A.;  Efremova,  L.  S.  1991.  Seed  prod- 
uction and  cone-feeding  insects  oiPinus  pumila  at 
Kamtchatka  peninsula:  aspects  of  coexistence.  In: 
Baranchikov,  Y.  N.;  Mattson,  W.  J.;  Hain,  F.  P.;  Payne, 
T.  L.,  eds.  Forest  insects  guilds:  patterns  of  interaction 
with  host  trees.  U.S.  Department  of  Agriculture,  Forest 
Service,  Gen.  Tech.  Rep.  NE-153:  316-320. 

Khomentovsky,  P.  A.;  Kazakov,  N.  V.;  Tchernyagina,  O.  A. 
1989.  Tundra-forest  zone  at  Kamtchatka:  problems  of 
conservation  and  management.  In:  Problems  of  nature 
management  in  taiga  zone  (collection  of  papers). 
Irkutsk:  30-46.  [In  Russian]. 

Khomentovsky,  P.  A.;  Khomentovskaya,  I.  G.  1990.  Geo- 
graphical variability  of  Pinus  pumila  seed  producing  at 
Kamchatka.  In:  Questions  of  Kamchatka  geography. 
Petropavlovsk-Kamchatsky.  No.  10:  47-55.  [In  Russian]. 


Krylov,  G.  V.;  Talantsev,  N.  K;  Kozakova,  N.  F.  1983. 
Siberian  pine  (Pinus  sibirica  Du  Tour).  Moscow.  216  p. 
[In  Russian]. 

Kunitsin,  L.  F.  1963.  A  pattern  of  Kamtchatka  nature 
regionalization.  In:  Nature  conditions  and  regionaliza- 
tion  of  Kamtchatka  district.  Moscow:  7-26.  [In  Russian]. 

Lukitcheva,  A.  N.  1964.  Distribution  areas  of  Asia  plants 
(a  map).  In:  World  geographical  atlas.  Moscow.  112  p. 

Malaeva,  E.  M.  1967.  Kamtchatka  vegetation  develop- 
ment in  Pliocene-Pleistocene.  In:  Development  of  Sibe- 
rian and  far-eastern  vegetation  in  Quaternary  (collec- 
tion of  papers).  Moscow:  78-170.  [In  Russian]. 

Pravdin,  L.  F.;  Iroshnikov,  A.  I.  1982.  Genetics  of  Pinus 
sibirica  Du  Tour,  P.  koraiensis  Sieb.  et  Zucc.  and  P. 
pumila  Regel.  Annales  Forestales.  9(3):  79-123. 

Rush,  V.  A.  1974.  Biochemical  characteristics  of  five- 
needle  pines  seeds.  In:  Biology  of  conifer  seed  repro- 
duction in  western  Siberia  (collection  of  papers). 
Novosibirsk:  180-184.  [In  Russian]. 

Sokolov,  I.  A.  1973.  Volcanism  and  soils  formation 
(with  Kamtchatka  as  an  example).  Moscow.  224  p. 
[In  Russian]. 

Sotchava,  V.  B.;  Lukitcheva,  A.  N.  1953.  On  geography 
of  dwarf  stone  pine.  Reports  of  the  USSR  Academy  of 
Sciences.  90(6):  1163-1166.  [In  Russian]. 

Tchelebaeva,  A.  I.;  Shantser,  A.  E.;  Egorova,  I.  A.;  Lupikina,  E. 
G.  1974.  Cenozoic  deposits  of  the  Kuril-Kamtchatka 
area.  In:  Kamtchatka,  Kuril  and  Commander  islands 
(series  of  History  of  the  relief  development.)  Moscow: 
31-57.  [In  Russian]. 

Tikhomirov,  B.  A.  1949.  Dwarf  stone  pine  {Pinus  pumila 
(Pall.)  Rgl),  its  biology  and  utilization.  MOIP  series, 
botan.  division,  Moscow.  14(6):  106  p.  [In  Russian]. 

Tstcherbakova,  E.  J.  1964.  An  average  of  maximal  decade 
snow  depths  (a  map).  In:  World  geographical  atlas. 
Moscow.  220  p. 


77 


NEEDLE  LONGEVITY  AND 
PHOTOSYNTHETIC  PERFORMANCE  IN 
CEMBRAN  PINE  AND  NORWAY  SPRUCE 
GROWING  ON  THE  NORTH-  AND  EAST- 
FACING  SLOPES  AT  THE  TIMBERLINE 
OF  STILLBERG  IN  THE  SWISS  ALPS 

Takayoshi  Koike 
Rudolf  Hasler 
Hans  Item 


Abstract — The  number  of  shoots  in  each  age  class  and  the  pho- 
tosynthesis of  the  shoots  were  measured  on  cembran  pine  {Pinus 
cembra)  and  Norway  spruce  (Picea  abies)  growing  on  north-  and 
east-facing  slopes  at  2,185  m  above  sea  level.  Needle  longevity 
on  the  north-facing  slope  exceeded  needle  longevity  on  the  east- 
facing  slope  by  1  to  3  years  in  both  species.  A  negative  correla- 
tion between  maximum  photosynthesis  and  needle  lifespan  was 
foimd.  However,  there  was  no  physiological  difference  in  the 
photosynthesis  of  trees  on  the  respective  slopes. 


In  the  Swiss  Alps,  cembran  pine  (Pinus  cembra)  grov^rs 
around  timberline  level  and  Nonvay  spruce  (Picea  abies) 
from  valleys  to  mountainsides  (Kuoch  and  Amiet  1970). 
At  the  subalpine  timberline,  the  growth  and  biomass  in- 
crease of  trees  are  strongly  influenced  by  the  microsite 
performance  and  microclimate  (Livingston  and  Black 
1987;  Schonenberger  and  Frey  1988;  Tvu-ner  and  others 
1982).  Using  the  monitoring  experiment  on  photosyn- 
thetic  production  as  a  basis,  the  biomass  increase  of 
mountain  pine  (Pinus  mugo)  growdng  at  the  timberline 
was  found  to  be  regulated  by  the  soil  temperature  and  the 
net  radiation  (Hasler  1982).  However,  physiological  char- 
acteristics of  trees  were  not  revealed.  Is  the  turnover  rate 
of  needles  related  to  the  tree  size?  What  about  the  physi- 
ological adaption  of  needles  to  the  microclimate? 

In  this  study,  these  questions,  needle  lifespan  and  the 
age-related  photosynthesis  of  cembran  pine  planted  on 
north-  and  east-facing  slopes,  were  measured.  We  present 
the  physiological  parameters  of  needle  photosynthesis. 
Anatomical  traits  of  needles  were  investigated  with  refer- 
ence to  the  estimation  of  the  intercellular  CO^  concentra- 
tions (C.)  (Terashima  1992).  We  discuss  the  trade-off  rela- 
tion between  the  needle  lifespan  and  photosynthetic  rates 
influenced  by  the  microclimate. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Takayoshi  Koike,  Rudolf  Hasler,  and  Hans  Item  are  members  of  the 
"Ecophysiology"  group,  Swiss  Federal  Institute  for  Forest,  Snow  and  Land- 
scape Research,  CH-8903  Birmensdorf,  Switzerland.  Present  address  of 
T.  Koike:  Forest  Products  Research  Institute,  Sapporo  062,  Japan. 


MATERIALS  AND  METHODS 

At  the  study  site  (Stillberg  in  the  Swiss  Alps),  on  the 
north-facing  slope,  the  yearly  net  radiation  surplus  was 
308  kWh/m^  lower,  and  the  root-zone  temperature  during 
grov^rth  period  was  2-4  °C  lower  than  on  the  east-facing 
slope  at  2,185  m  above  sea  level  (a.s.l.)  (Turner  and  others 
1982).  The  mean  soil  temperature  at  10  cm  on  the  north- 
and  east-facing  slopes  from  June  to  September  was  3.5 
and  4.5  °C,  respectively  (Koike  and  others,  submitted). 

Plant  Materials 

Plants  studied  were  cembran  pine,  (provenance  2,050  m 
a.s.l.)  and  Norway  spruce  (provensmce  1,960  m  a.s.l.), 
raised  from  seeds  in  the  nursery  at  Birmensdorf  ZH  (550  m 
a.s.l.).  The  cembran  pines  were  transplanted  in  1975, 
and  the  Norway  spruces  in  1962,  to  the  experimental  site 
(2,185  m  a.s.l.).  Mean  tree  height  as  of  1988  on  the  north- 
facing  slope  and  east-facing  slope  was  40  cm  and  80  cm  in 
cembran  pine  and  36  cm  and  45  cm  in  Norway  spruce,  re- 
spectively. For  measurements  of  photos3mthesis,  attached 
shoots  vdthin  the  even-aged  needles  were  used.  Younger 
shoots  were  removed  when  the  photosynthesis  of  older 
shoots  was  measured.  The  cut  shoot  ends  were  sealed 
with  vaseline. 

Measurements 

The  gas  exchange  rate  was  measured  with  a  thermo- 
electrically  controlled  chamber  (Mini-cuvette  Walz,  FRG), 
which  was  coated  with  aluminium  foil  in  order  to  fully 
use  diffuse  light  (Koike  and  others,  submitted).  Photosyn- 
thesis was  determined  v^dth  two  infrared  gas  analyzers 
(Binos,  Leybold-Heraeus,  FRG).  Dew  point  mirrors  (Walz, 
FRG)  were  used  to  measure  the  absolute  air  hmnidity. 
Needle  temperatures  were  monitored  by  a  0.1  mm  chromel- 
constantan  thermocouple.  The  PPFD  above  the  needles 
was  measured  with  a  GaAsp  photodiode  (Hamamatsu 
Gil  18,  J)  after  calibration  vsrith  a  LiCor  quanttmi  sensor 
(USA).  A  COj  dispensing  apparatus  (Walz,  FRG)  pro- 
duced different  COg  concentrations  in  the  chamber. 


78 


Figure  1 — Cross  section  of  a  needle  of  a  cembran 
pine  (right)  and  a  NoPA'ay  spruce  (left). 


Anatomy  of  needles  was  observed  with  a  microscope 
after  needles  were  sectioned  with  a  microtome  (Leitz). 
Total  needle  surface  area  of  the  used  twig  was  estimated 
by  the  modified  glass  bead  method  (Davies  and  Benecke 
1980).  The  number  of  specimens  for  the  census  of  needle 
longevity  was  more  than  700  shoots  of  10  trees.  Dry 
weight  of  the  needles  was  determined  after  48  hours  dry- 
ing at  80  °C.  Nitrogen  concentration  in  the  needles  was 
analyzed  with  a  CHN  analyzer  (Rapid,  USA). 

RESULTS 

Needle  Characteristics 

The  anatomy  of  needles  of  cembran  pine  and  Norway 
spruce  was  similar  to  the  "homobaric"  leaves  (fig.  1). 
There  was  no  extension  of  vascular  bundle  sheaths  into 
the  mesophyll.  Needle  lifespans  of  cembran  pine  and 
Norway  spruce  were  9  and  12  years  for  the  north-facing 
slope  and  8  and  9  years  for  the  east-facing  slope,  respec- 
tively (fig.  2).  The  proportion  of  older  needles  of  both  spe- 
cies was  larger  on  the  north-facing  slope  than  on  the  east- 
facing  slope.  In  Norway  spruce,  the  frequency  of  younger 
needles  on  the  north-facing  slope  was  smaller  than  that 
on  the  east-facing  slope.  A  negative  correlation  between 
the  Ught-saturated  photosynthesis  and  needle  longevity 
was  found  in  both  species  (fig.  3). 


Physiological  Parameters 

There  was  a  positive  correlation  between  the  stomatal 
conductance  and  the  light-saturated  photosjoithesis  in 
both  species  (fig.  4).  No  difference  between  the  north-  and 
east-facing  slopes  was  foimd.  The  quantum  yield  and  car- 
boxylation  efficiency  (CE)  of  both  species  on  the  north- 
facing  slope  tended  to  be  lower  than  of  those  on  the  east- 
facing  slope  (table  1). 

The  maximimi  photo sjTithe sis  at  light  and  CO^  satura- 
tion (Pj^^)  of  both  species  on  the  north-facing  slope  was 
lower  than  that  on  the  east-facing  slope  (P  <  0.05).  The 
CE  of  cembran  pine  was  sHghtly  higher  than  that  of 
Norway  spruce.  The  nitrogen  concentration  in  needles 
of  cembran  pine  was  higher  than  in  needles  of  Norway 
spruce.  However,  there  was  no  statistical  difference  in 
both  species  between  the  slopes. 

DISCUSSION 

The  estimation  of    is  based  on  the  uniform  response 
of  the  stomata  of  a  leaf  (Terashima  1992).  The  anatomy 
of  needles  of  cembran  pine  and  Norway  spruce  was  simi- 
lar to  the  homobaric  leaf.  Non-uniform  stomatal  aperture 
in  needles  of  loblolly  pine  was  observed  only  when  needles 


Figure  2 — Frequency  of  needle  age  in  percent  of  Figure  3 — Photosynthesis  in  relation  to  needle  age 

total  needles  of  cembran  pines  (right)  and  Norway  in  cembran  pines  (left)  and  Norway  spruces  (right) 

spruces  (left)  growing  on  east-facing  and  north-  growing  on  east-facing  and  north-facing  slopes  at 

facing  slopes  at  timberline.  timberline. 


79 


8 


sz 

c 

w 
o 

o 
sz 


'E 

-54 

E 


1 

1 

•  North-facing  slope 

o 

uusi-Taciny  siop6 

1 

1 

0  40         80  0  40  80 

Stomotal  conductance (mmolrfi^s^) 

Figure  4 — Photosynthesis  in  relation  to  stomatal 
conductance  in  cembran  pines  (left)  and  Norway 
spruces  (right)  growing  on  east-facing  and  north- 
facing  slopes  at  timberline. 


were  treated  with  abscisic  acid,  not  with  the  low  tempera- 
ture in  roots  (Day  and  others  1991).  According  to  these 

facts,  we  can  estimate  the  CE  calculated  with  C.  values. 

1 

The  P^^  from  needles  of  the  north-facing  slope  was 
smaller  than  of  those  of  the  east-facing  slope,  while  needle 
lifespan  of  the  north-facing  slope  increased.  Many  physi- 
ological parameters  in  needles  from  the  north-facing  slope 
were  almost  the  same  as  those  on  the  east-facing  slope. 
However,  slightly  lower  P^^  in  both  species  from  the 
north-facing  slope  may  be  attributed  to  the  lower  soil 
temperature  (Day  and  others  1991;  Hasler  1982,  Turner 
and  others  1982).  The  lower  P     of  needles  from  the 

max 

north-facing  slope  may  be  compensated  for  by  the  longer 
lifespan  of  needles.  Tree  growth  on  the  north-facing  slope 
may  be  retarded  by  the  shortage  of  radiation  and  low  soil 
temperatures. 

ACKNOWLEDGMENTS 

We  thank  Dr.  H.  Turner  for  providing  the  climate 
data  of  the  Stillberg  site.  Prof.  W.  K.  Smith  for  collecting 
references,  and  Mrs.  M.  J.  Sieber  for  editing  the  English 
version  of  this  manuscript. 


REFERENCES 

Davies,  C.  E.;  Benecke,  U.  1980.  Fluidized  bed  coating 
of  conifer  needles  with  glass  beads  for  determination 
of  leaf  surface  area.  Forest  Science.  26(1):  29-32. 

Day,  T.  A.;  Heckathorn,  S.  A.;  DeLucia,  E.  H.  1991.  Limi- 
tations of  photosynthesis  in  Pinus  taeda  L.  (loblolly 
pine)  at  low  soil  temperatures.  Plant  Physiology.  96: 
1246-1254. 

Hasler,  R.  1982.  Net  photosynthesis  and  transpiration 
of  Pinus  montana  on  east  and  north  facing  slopes  at 
alpine  timberline.  Oecologia.  54:  14-22. 

Koike,  T.;  Hasler,  R.;  Matyssek,  R.;  Item,  H.  Seasonal 
changes  in  the  photosynthetic  capacity  of  Larix  decidua 
and  Pinus  cembra  planted  on  contrasting  slopes  at  the 
timberline  at  Stillberg  (Davos),  eastern  Switzerland. 
Trees,  submitted. 

Kuoch,  R.;  Amiet,  R.  1970.  Die  Verjiingung  im  Bereich  der 
oberen  Waldgrenze  der  Alpen.  Eidgenossische  Anstalt 
fiir  das  forstliche  Versuchswesen,  Mitteilungen.  46: 
159-328. 

Livingston,  N.  J.;  Black,  T.  A.  1987.  Stomatal  characteris- 
tics and  transpiration  of  three  species  of  conifer  seed- 
lings planted  on  a  high  elevation  south-facing  clear-cut. 
Canadian  Journal  of  Forest  Research.  17:  1273-1282. 

Schonenberger,  W.;  Frey,  W.  1988.  Untersuchungen  ziar 
Okologie  imd  Technik  der  Hochlagenaufiforstimg. 
Forschungsergebnisse  aus  dem  Lawinenanrissgebiet 
Stillberg.  Schweizerische  Zeitschrift  fur  Forstwesen. 
139:  735-820. 

Terashima,  L  1992.  Anatomy  of  non-uniform  leaf  photo- 
synthesis. Photos)aithesis  Research.  31(3):  195-212. 

Turner,  H.;  Hasler,  R.;  Schonenberger,  W.  1982.  Contrast- 
ing microenvironments  and  their  effects  on  carbon  up- 
take and  allocation  by  yoimg  conifers  near  alpine  tree- 
line  in  Switzerland.  In:  Waring,  R.  H.:  Carbon  uptake 
and  allocation  in  subalpine  ecosystems  as  a  key  to  man- 
agement: Proceedings  of  the  lUFRO  Workshop, 
Pl.07-00,  Corvalhs,  OR:  Oregon  State  University: 
22-30. 


Table  1 — Physiological  parameters  of  cembran  pine  and  Norway  spruce  growing  on  an  east-  and  a  north-facing  slope  of  an 
avalanche  gully  at  Stillberg  (2,185  m  a.s.l.) 


Parameter 


Cembran  pine 


E  slope 


N  slope 


Norway  spruce 


E  slope 


N  slope 


Quantum  yield  0.014  ±0.004 
(mol  COj'mol/quanta) 

Carboxylation  efficiency  0.024  ±  0.009 
(^mol•m-2•^ba^^' 


(mol»m-2»s-^) 


8.30  ±  1.51 


Nitrogen  concentration 
(mg»g-^)  12.26  ±  0.90 


0.009  ±  0.002 

0.022  ±  0.009 

5.38  ±  0.77 

13.27  ±  0.91 


ns 


ns 


(1) 


ns 


0.012  ±  0.006       0.006  ±  0.002 


0.017  ±  0.002       0.018  ±  0.008 


6.87  ±  1.61 


10.43  ±  1.87 


4.93  ±0.12 


9.26  ±  1.22 


ns 


ns 


(1) 


ns 


'P<0.05. 


80 


THE  BROAD-LEAVED  KOREAN  PINE 
FOREST  IN  CHINA 

Luo  Ju  Chun 


Abstract — Korean  pine  (Pinus  koraiensis)  is  a  tree  species  of 
high  economic  value.  Its  natural  distribution  area,  the  character- 
istics of  distribution  in  the  broad-leaved  Korean  pine  forest  re- 
gion of  China,  the  biological  and  ecological  characteristics  of  Ko- 
rean pine,  the  forest  types  of  Korean  pine,  and  the  succession 
pattern  of  the  Korean  pine  forest  communities  have  been  studied. 
The  results  are  significant  to  conservation  and  sustaining  use  of 
Korean  pine. 


Korean  pine  (Pinus  koraiensis)  is  well  known  for  its 
high-quality  timber  that  can  be  used  for  many  different 
purposes.  Undoubtedly,  Korean  pine  is  one  of  the  most 
important  tree  species  in  Chinese  forests. 

Except  for  a  few  pure  Korean  pine  forests,  the  species 
often  occurs  in  the  "conifer-broad-leaved  tree  mixed-forest 
communities"  vnth  many  species  of  deciduous  broad- 
leaved  trees  and  other  conifers,  which  is  the  most  repre- 
sentative vegetation  formation  in  the  northeastern  humid 
zone  in  China. 

NATURAL  DISTRIBUTION  AREA 

Korean  pine — a  relict  species  from  the  Tertiary — 
mainly  occurs  in  northeast  Asia  and  in  a  narrow  zone 
along  the  west  coast  of  the  Japan  Sea.  This  zone  includes 
the  southern  part  of  eastern  Russia,  the  eastern  part  of 
northeast  China,  North  Korea,  and  the  center  of  Honshu 
in  Japan.  The  main  natural  range  of  Korean  pine,  how- 
ever, is  in  China.  Korean  pine  is  a  representative  species 
in  Dahurian  flora,  and  also  is  the  main  forest  tree  species 
in  the  eastern  momitains  of  northeast  China. 

In  China,  the  broad-leaved  Korean  pine  forest  is  found 
from  40°  15'  N.  to  50°20'  N.  and  from  126°  E.  to  135°  E., 
east  of  Song-nen  plain  and  north  of  Song-liao  plain.  In 
the  south  it  reaches  Dan  dong,  in  the  north  Hei-Ho 
coimty.  The  whole  sirea  takes  a  crescent  shape  covering 
Lesser  Xingan  Ling,  Wan  Da  Mountain,  Zhang  Guang  Cai 
Ling,  Lao  Ye  Ling,  and  Chang  Bai  Mountain.  Most  of 
these  mountains  are  trending  from  northeast  to  south- 
west. They  are  characterized  by  gentle  topography  and 
low  elevation,  not  higher  than  1,300  m.  There  are  many 
different  tree  species,  including  many  quite  valuable 
broad-leaved  trees.  Their  exploitation  began  more  than 
100  years  ago.  Their  timber  output  still  makes  up  one- 
fourth  of  total  timber  production  in  China. 

The  forest  region,  which  is  influenced  by  the  Japan  Sea, 
is  characterized  by  temperate-zone  monsoon  climate.  The 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Luo  Ju  Chun  is  Professor  of  Ecology  and  Forest  Management,  Forestry 
Resource  College,  Forestry  University,  Beijing,  China. 


warm  and  moist  climate  is  favorable  to  growth  of  Korean 
pine.  Annual  average  temperature  is  0-6  °C;  the  growing 
season  is  about  4-5  months.  Temperatvu"e  svun  (>10  °C) 
is  about  2,500-3,200  °C.  At  higher  latitude,  temperature 
is  low,  vyinter  is  long,  and  simimer  short.  Winter  covers 
5  months.  January  is  the  coldest  month;  the  absolute 
minimum  temperature  is  -40  to  -30  °C.  In  July,  which  is 
the  warmest  month,  mean  temperature  reaches  20-24  °C. 
Winter  is  characterized  by  thick  snow  cover  and  mean 
temperatures  ranging  from  -28  to  -14  °C.  Annual  pre- 
cipitation amoimts  to  600-1,100  mm  £ind  decreases  from 
south  to  north;  maximum  is  in  summer  (June  to  August). 
The  annual  relative  hiunidity  is  about  60-70  percent. 
Hiunidity  and  temperature  are  favorable  to  growth  of 
Korean  pine  and  other  trees. 

The  soil  in  Korean  pine  forests  is  dark  brown  forest  soil 
vfith  a  thick  hvunus  layer,  and  it  is  very  fertile. 

CHARACTERISTICS  OF 
VEGETATION  DISTRIBUTION 

Horizontal  Distribution 

Because  of  the  wide  range  of  Korean  pine  forest,  the  cli- 
mate is  quite  different  from  north  to  south  in  the  distribu- 
tion area.  Consequently,  some  differences  in  flora  and 
plant  commvuiities  become  apparent.  The  distribution 
area  can  be  divided  into  two  subzones:  northern  and 
southern  temperate.  The  demarcation  is  from  Don-g  Nin 
in  the  eastern  part  of  Jing  Bo  Lake  to  Ji  Lin  district  until 
its  west  boiindary;  it  seems  to  nm  along  the  parallel  of 
44  °N.  The  differences  in  natiiral  conditions  and  vegeta- 
tion can  be  described  as  follows: 

1.  In  the  northern  subzone,  climate  is  cold,  the  growing 
season  is  only  about  100-120  days,  and  annual  rainfall 
£imounts  to  500-700  mm;  the  southern  subzone  is  charac- 
terized by  a  warm  and  hvunid  climate  and  a  growing  sea- 
son of  130-150  days.  Annual  precipitation  is  about  700- 
1,100  mm. 

2.  In  the  northern  subzone,  more  conifers  typical  of  the 
cold-temperate  zone  (for  example,  Picea  koraiensis,  P. 
jezoensis,  and  Abies  nephrolepis)  occur  in  Korean  pine 
forests.  Pinus  pumila  can  be  found  in  the  subalpine.  In 
addition,  Quercus  mongolica  is  more  common  on  expo- 
siu-es  to  Bun.  Larix  gmelinii  forests,  mixed  with  Betula 
platyphylla,  occur  in  the  lowlands.  Vaccinium  vitis- 
idaea,  Ledum  palustre,  Vaccinium  uliginosum,  Betula 
fructicosa — representatives  of  Dahurican  flora — are  typi- 
cal of  larch  forest  undergrowth.  In  the  southern  subzone, 
there  are  more  representatives  of  Chang  Bai  flora,  such  as 
Picea  jezoensis  var.  komarovii,  Abies  holophylla,  and 
Larix  olgensis. 


81 


3.  In  the  southern  subzone  there  are  more  plant  species 
than  in  the  north.  The  forest  often  consists  of  30-40  woody 
species  forming  three  layers.  In  the  southern  subzone 
there  are  twice  as  many  broad-leaved  tree  species  in 
Korean  pine  forests  than  in  the  northern  subzone;  the 
genus  Acer,  for  instance,  is  represented  by  only  five  spe- 
cies in  the  north,  compared  to  12  species  in  the  south. 
The  percentage  of  valuable  broad-leaved  trees  (such  as 
Franus  mandshurica,  Juglans  mandshurica,  Phello- 
dendron  amurense,  Tilia  amurensis,  Ulmus  propingua) 
in  Korean  pine  forests  is  higher  in  the  southern  than  in 
the  northern  subzone.  These  broad-leaved  trees  usually 
grow  together  with  Betula  costata  and  Acer  mono  and  rep- 
resent more  than  50  percent  of  the  species  of  southern 
Korean  pine  forests.  Moreover,  Carpinus  cordata  and 
Fraxinus  rhynchophylla  are  very  commom  in  the  south, 
but  occur  rarely  in  the  north.  In  addition,  there  are  more 
than  10  vine  plant  species  in  the  southern  subzone,  but 
only  five  or  six  in  the  north. 


1400  -r 


0-^ 


Figure  1 — Vertical  distribution  pattern  of  veg- 
etation in  Lesser  Xin  An  Ling  (southern  expo- 
sures): 1a,  River  head-valley-conifer  forest  with 
larch,  spruce,  fir;  1b,  Bank  depression,  broad- 
leaved  forest  with  Ulmus  propinqua,  Fraxinus 
mandshurica,  Pliellodendron  amurense;  2,  Oak 
forest  {Quercus  mongolica)  mixed  with  Betula 
dahurica,  Tilia  mandshurica;  3,  Broad-leaved 
Korean  pine  forest  mixed  with  spruce,  fir, 
Betula  costata,  Tilia  amurensis,  Ulmus 
laciniata;  4,  Spruce-fir  forest  belt;  5,  Subalpine 
Betula  ermanii-Pinus  pumila  elfin  forest  belt. 


Vertical  Distribution 

The  vertical  distribution  pattern  of  vegetation  in  Lesser 
Xin  An  Ling  (fig.  1)  is  a  typical  example  of  the  northern 
subzone: 

1.  River  valley  forest:  It  is  composed  of  Picea  jezoensis, 
Abies  nephrolepis,  Larix  gmelinii,  Juglans  mandshurica, 
Ulmus  propinqua,  Populus  ussuriensis,  and  other  species. 

2.  Oak  {Quercus  mongolica)  broad-leaved  forest  belt 
(150-300  m):  It  is  secondary  forest  that  established  after 
destruction  of  virgin  conifer-broad-leaved  mixed  forest, 
and  it  occurs  usually  on  shallow  soils  on  s\in-exposed 
slopes. 

3.  Broad-leaved  Korean  pine  forest  belt  (250-650  m): 
Korean  pine  is  the  dominant  species;  spruces  and  firs  oc- 
casionally occur;  the  broad-leaved  trees  make  up  less  than 
30  percent  of  the  species.  The  volume  of  standing  timber 
is  the  highest  (300-600  m^/ha)  in  the  northern  forest  re- 
gion. Korean  pine  grows  on  exposures  to  south  and  south- 
west, on  mountain  ridges,  and  on  hill  tops. 

4.  Spruce-fir  forest  belt  (650-1,000  m):  Picea  korai- 
ensis,  Picea  jezoensis,  and  Abies  nephrolepis  are  the  main 
tree  species. 

5.  Subalpine  moss-elfin  forest  belt  (1,000-1,080  m): 
The  climate  is  cold  and  humid;  wind  is  strong.  The  elfin 
forest  is  composed  of  Betula  ermanii.  This  species  exhib- 
its dwarf  growth  forms  not  taller  than  4-6  m.  Pinus 
pumila  grows  in  the  forest  understory.  The  forest  floor 
is  covered  mainly  by  moss. 

The  vertical  distribution  pattern  of  vegetation  in  the 
southern  subzone  can  be  illustrated  by  the  situation  of 
Chang  Bai  Mountain  (fig.  2): 

1.  Broad-leaved  forest  belt  (250-500  m  a.s.l.):  It  is 
formed  by  Quercus  mongolica,  Betula  platyphylla,  Betula 
dahurica,  and  Populus  davidiana.  It  is  a  typical  secon- 
dary forest  following  destruction  of  broad-leaved  Korean 
pine  forest.  Locally,  basswood,  maple,  elm,  walnut,  and 
corktree  form  mixed  forests.  At  lower  elevations,  these 
broad-leaved  forests  are  scattered. 

2.  Broad-leaved  Korean  pine  forest  belt  (500-1,200  m): 
Korean  pine  is  the  main  species,  and  is  mixed  with  vari- 
ous broad-leaved  trees:  Betula  costata,  Carpinus  costata, 
Ulmus  laciniata,  Acer  mono,  Tilia  amurensis,  Quercus 
mongolica,  especially  the  valuable  broad-leaved  tree  spe- 
cies Ulmus  propinqua,  Phellodendron  amurense,  Juglans 
mandshurica,  Fraxinus  mandshurica,  and,  at  valley  bot- 
toms, some  spruce  and  fir  can  be  found. 

3.  Spruce-fir  forest  belt  (1,200-1,800  m):  It  is  formed 
by  Picea  jezoensis,  Picea  koraiensis,  and  Abies  nephrolepis, 
with  additional  Betula  ermanii  and  Larix  olgensis  in  the 
upper  part  of  the  belt,  or  with  Korean  pine,  Betula 
costata,  Ulmus  laciniata,  and  Tilia  amurensis  in  its  lower 
part. 

4.  Subalpine  Betula  ermanii  mossy  elfin  forest  belt  and 
subalpine  meadow  belt  (1,800-2,100  m):  These  belts  are 
characterized  by  cold  and  windy  climate,  steep  topogra- 
phy, and  shallow  soils.  Betula  ermanii  forests  alternate 
with  vast  meadow  areas. 

5.  Alpine  tundra  belt  (>2,100  m):  Vegetation  consists 
of  different  dwarf  shrub  species,  mosses,  and  lichens. 


82 


3000  -r 


2500 


2000  - 


(0 

5 1600  - 
0) 


1000  - 


500  - 


0-^ 


Figure  2 — Vertical  zonal  spectrum  of  vegetation 
in  Chang  Bai  Mountain:  1 ,  Secondary  deciduous 
broad-leaved  forest;  2,  Conifer-broad-leaved 
mixed  forest  (broad-leaved  Korean  pine  forest); 
3,  Coniferous  forest  belt:  3a,  Spruce-fir-Korean 
pine  forest  subbelt;  3b.  Spruce-fir  coniferous  for- 
est subbelt;  4,  Subalpine  Betula  ermanii-e\i\n  for- 
est belt,  and  subalpine  meadow  belt;  5,  Alpine 
tundra  belt. 


As  far  as  we  know,  the  higher  the  latitude  the  lower  is 
the  upper  limit  of  Korean  pine.  In  the  southern  subzone 
Korean  pine  occurs  up  to  1,200  m  a.s.l.;  in  the  north  its 
uppermost  altitudinal  limit  is  located  at  about  650  m. 

BIOLOGICAL  AND  ECOLOGICAL 
CHARACTERISTICS 

Korean  pine  is  a  long-lived  species.  Its  lifespan  is  300- 
400  years,  sometimes  500  years.  Tree  height  usually 
reaches  30  m.  The  diameter  can  reach  1  m.  Korean  pine 
is  one  of  the  valuable  sawtimber  trees  in  the  world.  It 
grows  slowly  in  the  first  5-8  years.  After  the  age  of  10  its 
growth  rate  increases  rapidly.  Its  diameter  growth  begins 
to  accelerate  at  the  age  of  12.  The  height  growth  becomes 
more  rapid  after  16-18  years.  Accelerated  growth  contin- 
ues vmtil  the  age  of  30.  Korean  pine  is  a  tree  species 
growing  fastest  at  middle  age.  In  general,  growth  rate  of 
Korean  pine  increases  earlier  in  plantations  than  in  natu- 
ral forests.  In  clear  cuttings,  the  radial  growth  of  8-year- 
old  trees  is  2.6  times  higher  than  in  nattu-al  forests,  and 
height  growth  is  twice  as  much  as  in  trees  at  the  same 
age  that  became  naturally  established. 


Secondary  growth  of  Korean  pine  is  common  in  autiunn, 
but  it  appears  mainly  in  trees  less  than  10  years  old. 
After  80  years  growth,  Korean  pine  reaches  a  height  of 
18-20  m.  At  this  age  the  tree  top  often  begins  to  fork  and 
bear  cones.  Obviously,  the  forking  is  closely  related  to 
cone  production. 

Korean  pine  has  shallow  roots.  Its  taproot  is  imderde- 
veloped  and  degenerates  easily.  Thus,  seed  trees  left  on 
clear  cuttings  are  sensitive  to  windbreak. 

Most  Korean  pine  trees  begin  to  produce  cones  at  the 
age  of  80-100  years.  Nevertheless,  cones  can  also  be 
found  on  30-year-old  trees.  In  plantations,  however, 
even  20-year-old  trees  are  able  to  bear  cones. 

There  are  two  ecot5T3es  of  Korean  pine:  one  is  the 
leptodermis  form  displaying  thin  bark  with  small  and 
shallow  scale  or  long  strip  lobes.  It  forks  less,  grows 
faster  in  height,  and  produces  better  timber.  The  other 
is  the  pachidermis  form  characterized  by  deep  and  large, 
long,  square-shape  lobes  on  the  tree  trunk.  It  forks  more, 
grows  slower  in  height,  and  the  timber  quality  is  not  as 
good  as  in  the  leptodermis  type.  The  leptodermis  type 
should  be  used  for  afforestation. 

Korean  pine  belongs  to  thermophile  species;  its  vegeta- 
tive growth  requires  6-7  °C  of  the  lowest  average  tempera- 
tiu-e  and  the  most  suitable  temperature  is  14-16  °C  in  the 
yoimg  stage.  The  relative  moistvu^e  it  demands  is  70  per- 
cent and,  during  the  period  of  most  rapid  growing,  rela- 
tive hvunidity  of  about  70  percent  is  required.  Korean 
pine  is  intolerant  to  high  temperature  during  vegetative 
growth.  If  average  temperature  exceeds  15-16  °C,  height 
growth  will  decrease  or  even  stop.  On  the  other  hand,  it 
will  not  be  affected  by  winter  temperatiires  as  low  as 
-50  °C. 

At  seedling  stage  a  little  shading  is  favorable  to  Korean 
pine,  althovigh  it  can  tolerate  full  siinlight.  Its  light  re- 
quirement increases  with  age.  While  3  years  old,  seed- 
lings require  60-70  percent  of  incoming  solar  radiation. 
After  the  fifth  year,  young  trees  become  totally  shade 
intolerant. 

Korean  pine  grows  best  on  thick,  moist,  fertile,  and  well- 
drained  soils  with  pH  5-6,  but  it  can  also  exist  on  shallow 
dry  soil,  poor  in  nutrients.  It  grows  badly  on  soils  rich  in 
clay.  However,  Korean  pine  requires  appropriate  soil  mois- 
ture. In  bogs  and  on  sites  with  stagnant  water,  growth  is 
hampered,  and  Korean  pine  is  rarely  seen  there. 

PRINCIPAL  FOREST  TYPES 

Mu  Dan  Jiang  forest  region,  the  distributive  center  of 
Korean  pine,  may  serve  as  an  example  to  classify  Korean 
pine  forest  into  the  following  forest  types: 

Steep-Slope  Carex  calltnichos-Korean  Pine 
Forest — This  type  occurs  on  simny  exposures  or  narrow 
ridges  at  an  altitude  of  800-1,000  m.  These  usually  are 
poor  sites.  The  conmiunity  can  be  divided  into  two  layers. 
The  forest  canopy  is  formed  by  Korean  pine  (90  percent) 
and  Quercus  mongolica  or  Tilia  amurensis  (10  percent), 
which  have  a  higher  crown  density  than  Korean  pine.  In 
the  second  layer  there  is  some  Korean  pine,  spruce,  fir, 
and  linden.  The  growing  stock  is  400-500  m^/ha.  The 
dominant  species  on  the  forest  floor  are  Carex  calltnichos 
and  C.  ussuriensis. 


83 


CoryluS'Carex-Korean  Pine  Forest— This  type  is 
found  on  mountains  with  gentle  slopes  at  altitudes  of 
700-900  m.  As  to  favorability  to  tree  growth,  these  sites 
can  be  considered  intermediate.  Korean  pine  prevails 
(60-70  percent)  followed  by  Tilia  amurensis  (20  percent), 
and  some  Abies  nephrolepis,  Picea  jezoensis,  Ulmus 
propinqua,  Acer  mono,  and  Betula  costata.  The  growing 
stock  is  about  400-500  m^/ha.  There  are  more  tree  species 
associated,  such  as  Acer  sp.  There  is  more  undergrowth, 
and  the  main  species  is  Corylus  mandshurica.  On  the  for- 
est floor  about  30  species  are  to  be  found.  Carex  sidero- 
sticta  and  ferns  are  the  most  common  plants. 

Fern-Spruce-Fir-Korean  Pine  Forest — This  type 
occurs  on  the  lower  gentle  slopes  (inclination  about  10°) 
of  the  mountains  or  along  side  streams  at  an  altitude  of 
500-800  m.  The  community  construction  is  complex.  The 
canopy  layer  is  formed  by  Korean  pine  and  some  Populus 
ussuriensis  and  Picea  jezoensis.  The  second  layer  is  com- 
posed of  Picea  jezoensis  and  P.  koraiensis  (50  percent), 
Abies  nephrolepis  (20  percent),  and  Tilia  amurensis 
(20  percent).  Other  species  are  Betula  costata,  Ulmus 
laciniata,  Acer  mono,  Phellodendron  amurense,  Juglans 
mandshurica,  Fraxinus  mandshurica,  and  many  kinds  of 
maple.  Because  of  better  site  conditions,  growing  stock  is 
higher  (250-350  m^/ha  for  the  first  story  and  200  m^/ha  for 
the  second  story).  The  undergrowth  is  abundant  with 
over  20  species;  about  30  species  occur  on  the  forest  floor. 
Ferns  are  dominant  (such  as  Dryopteris). 

Fern-Moss  Korean  Pine  Forest — This  type  is  found 
in  the  lower  part  of  mountains  or  in  broad  valleys;  slope 
inclination  is  0-5°.  Two  stories  can  clearly  be  distin- 
guished: Pinus  koraiensis  (80  percent)  and  a  few  Picea 
jezoensis,  Fraxinus  mandshurica,  Ulmus  propinqua,  Acer 
mono,  and  Tilia  amurensis  form  the  forest  canopy.  In  the 
second  story  there  is  more  Picea  jezoensis,  Betula  costata, 
Acer  mono,  Abies  nephrolepis,  Tilia  amurensis,  and  a  few 
Juglans  mandshurica,  Fraxinus  mandshurica,  Ulmus 
propinqua,  Betula  platyphylla,  and  many  kinds  of  maple. 


So  the  tree  species  are  very  abundant.  This  forest  type 
is  typical  of  the  sites  most  favorable  to  tree  growth.  The 
growing  stock  can  reach  600  m^/ha.  The  imdergrowth 
mainly  consists  of  Lonicera  sp.,  Viburnum  sp.,  and  Sor- 
baria,  and  is  well  developed.  Carex  and  multiple  fern  and 
luxuriant  moss  cover  are  common  throughout  the  area. 

COMMUNITY  SUCCESSION 
PATTERNS 

Two  succession  patterns  for  the  broad-leaved  Korean 
pine  communities  are  obvious: 

Succession  Without  Disturbance — Natural  regen- 
eration is  poor  under  virgin  Korean  pine  forest  cover,  al- 
though a  comparatively  large  quantity  of  seedlings  may 
appear  after  a  seed  year.  However,  they  will  only  be  able 
to  grow  up  if  the  canopy  is  cut  thin.  Because  Korean  pine 
lives  longer  than  other  coniferous  and  broad-leaved  trees, 
and  tolerates  shade  for  longer  period,  it  tends  to  form 
multigeneration  and  multistoried  forests. 

Succession  After  Disturbance — ^After  destruction  of 
broad-leaved  Korean  pine  forest  the  succession  goes  on  as 
shown  by  the  following  sequences: 

Selective  cutting  -»  conifer-broad-leaved  mixed  forest 
in  which  main  elements  are  broad-leaved  trees.  Protec- 
tion    broadleaved-Korean  pine  forest. 

Clear  cutting  or  fire  ->  secondsiry  bare  land.  Protec- 
tion    grass  and  shrub  communities  conifer-hardwood 
mixed  forests  in  which  main  elements  are  hardwood  trees 
->  conifer-broad-leaved  mixed  forest  in  which  main  ele- 
ments are  coniferous  trees     broad-leaved-Korean  pine 
forest. 

To  promote  the  reestablishment  of  Korean  pine  forest, 
management  should  be  adjusted  to  this  succession 
process. 


84 


CLIMATES  WHERE  STONE  PINES 
GROW,  A  COMPARISON 

Tad  Weaver 


Abstract— While  stone  pine  climates  are  similar,  species 
adapted  to  relatively  moderate  climates  may  be  excluded  from 
the  ranges  of  congeners  by  more  severe  climates,  and  species 
with  longer  warm-moist  growing  seasons  are  probably  more  pro- 
ductive than  congeners.  Absolute  low/summer  average/absolute 
high  temperatures  for  stone  pines  listed  in  order  of  increasing 
absolute  low  temperature  are  Pinus  sibirica  (-55/13/37  °C), 
P.pumila  (-52/9/36  °C),  P.  koraiensis  (-42/11/36  °C),  P.  albi- 
caulis  (-34/9/29  °C),  and  P.  cembra  (-23/8/27  °C).  The  Walter 
drought  index  shows  little  stress  in  stone  pine  forests  despite 
large  differences  in  summer/winter  precipitation:  in  order  of 
increasing  summer  rainfall,  precipitation  is  P.  albicaulis  (102/ 
829mm),  P.  pumiVa  (142/264mm),  P.  sibirica  {187 /245mm), 
P.  cembra  (323/616mm),  P.  koraiensis  (394/242mm).  Estimated 
thawed-soil  growing  season  increases  from  P.  albicaulis  (4.5mo), 
through.  P.  pumila  (4.6mo),  P.  sibirica  (5.5mo),  and  P.  cembra 
(6.3mo)  to  P.  koraiensis  (7.8mo);  growing  seasons  of  the  first 
three  trees  could  be  shortened  by  drought. 


Stone  pines  grow  in  most  of  the  boreal  zone  (fig.l) 
(Fullard  and  Darby  1964;  Lanner  1990;  Mirov  1967). 
Pinus  pumila  and  P.  sibirica  occupy  vast  far-northern 
(50  to  70°  N.  latitude)  areas  in  northeastern  Asia;  P.  sibir- 
ica ranges  from  just  west  of  the  Urals  (55°  E.  longitude) 
halfway  to  the  Pacific  (115°  E.  longitude)  and  P.  pumila 
ranges  from  115°  E.  longitude  to  the  Pacific  (165°  E.  longi- 
tude). P.  koraiensis  grows  at  the  boreal-deciduous  forest 
transition  in  eastern  China  (120  to  150°  E.  longitude  and 
45  to  55°  N.  latitude  with  outliers  to  35°  N.  latitude  at 
high  altitude).  Paralleling  P.  koraiensis,  P.  albicaulis 
occupies  the  Rocky-Cascade-Sierra  Mountain  chains  of 
North  America  from  55°  N.  latitude  southward  to  45°  N. 
latitude  with  outhers  to  near  35°  N.  latitude.  Pinus 
cemba  appears  in  the  Alps  at  approximately  45°  N.  lati- 
tude. Glaciated  parts  of  the  conifer  forest  zone— both  in 
northern  Europe  and  North  America— lack  stone  pines 
(fig.l). 

Due  to  their  common  ancestry  and  common  boreal 
forest  habitat,  one  expects  the  trees  to  occupy  similar  cli- 
mates. This  paper  tests  that  hypothesis  by  comparing  cli- 
mates occupied  by  the  trees  with  respect  to  20factors  that 
may  be  important  in  determining  the  ranges  and  produc- 
ivities  of  the  species.  The  information  may  be  especially 
useful  for  predicting  the  success  of  introductions— from 
one  region  to  another— of  stone  pines,  their  tree  associates, 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Statxis  of  Our  Knowledge,  St.Moritz, 
Switzerland,  September5-ll,  1992. 

Tad  Weaver  is  Plant  Ecologist,  Biology  Department,  Montana  State 
University,  Bozeman,  MT  59717. 


and  to  some  degree,  even  their  herb,  crj^jtogam,  and  ani- 
mal associates.  Such  introductions  might  be  of  interest 
as  either  producers  or  pests. 

METHODS 

Climates  of  environments  dominated  by  closely  related 
stone  pines  (Lanner  1990;  Mirov  1967)  were  compared. 
The  comparisons  are  nonstatistical,  since  the  objective 
was  to  discuss  biological  responses  rather  than  the  cli- 
mates themselves. 

The  climates  were  characterized  by  choosing  three  to 
four  stone  pine  stands  at  which  temperature  and  precipi- 
tation were  regularly  measured  (a  complete  sample  of 
accessible  data);  summarizing  the  data  for  periods  of  10 
years  (P.  albicaulis  and  P.  cembra)  or  longer  (unspecified); 
and  calculating  means  for  selected  parameters.  The  sta- 
tions studied  are  Usted  here— from  west  to  east— each 
with  its  approximate  latitude,  longitude,  and  altitude  (m). 
Note  that  P.  albicaulis  and  P.  cembra  grow  significantly 
higher  than,  and  to  the  south  of,  the  Asian  pines.  Aster- 
isked Siberian  stations  lack  most  temperature  data.  Sci- 
entists who  helped  identify  weather  stations  in  stone  pine 
climates  are  listed  in  the  acknowledgments. 

Pinus  albicaulis  stations  (USDC  1961-80)  were  Crater 
Lake,  OR  (43°  N.,  122°  W.,  1,990m),  Ellery  Lake,  CA 
(38°  N.,  119°  W.,  2,940m),  Old  Glory,  BC  (49  °  N.,  119°  W., 
1,008m),  and  Kings  Hill,  MT  (46  °  N.,  110°  W.,  2,225m). 
Pinus  cembra  stations  (contributed  by  W.Tranquillini) 
were  Haggen  (47°  N.,  11°  E.,  1,800m),  Obergurgl  (47  °  N., 
11°  E.,  2,070m),  and  Patscherkofel  (48  °  N.,  11°  E., 
1,952m).   Pinus  sibirica  stations  (Muller  1982)  were 
Serov  (60°  N.,  61°  E.,  132m*),  Surgut  (61  °  N.,  73°  E., 
40m),  Kolpasevo  (58  °  N.,  83°  E.,  76m*),  Jenisejsk 
(58°  N.,  92°  E.,  78m*),  Krasnojarsk  (56  °  N.,  93°  E., 
15m),  Tura  (64  °  N.,  100°  E.,  130m*),  Irkutsk  (52  °  N., 
104°  E.,  468m),  and  Kirensk  (58  °  N.,  108°  E.,  256m). 
Pinus  pumila  stations  (Muller  1982)  were  Vitujsk  (64°  N., 
122°  E.,  107m*),  Jakutsk  (62  °  N.,  130°  E.,  100m*), 
Verchojansk  (68°  N.,  133°  E.,  137  m),  Ochotsk  (59°  N., 
143°  E.,  6  m),  Zyranka  (65°  N.,  151°  E.,  43m*), 
Petropavlovsk-Kamcatskij  (53°  N.,  159°  E.,  32m),  Apuka 
(65°  N.,  170°  W.,  10m*),  and  Anadyr  (65  °  N.,  178°  E., 
62m).  Pinus  koraiensis  stations  (contributed  by  Luo 
Ju  Chun  and  Zhao  Shidong)  were  AnTu  (43°  N.,  128°  E., 
591m),  Chargbei-ShenYang  (42  °  N.,  128°  E.,  738m)  and 
YiChun  (48°  N.,  129°  E.,  231m). 

Parameters  studied  describe  climates  with  respect  to 
temperature,  precipitation,  and  growing  season.  The  fol- 
lowing three  paragraphs  list  these  parameters  and  outline 
the  rationales  for  their  choice;  the  rationales  are  devel- 
oped further  in  Weaver  (1993)  and  in  the  following 
discussion. 


85 


B 

7^ — 

•1 

Figure  1 — Distribution  of  stone 
pines  in  relation  to  space,  conifer 
forests,  and  pleistocene  glaciation. 
North  America  and  Eurasia  are 
seen  from  a  polar  view.  Shading 
locates  the  conifer  zone  on  the  two 
continents  (Fullard  and  Darby 
1964).  Areas  within  the  dotted  lines 
received  pleistocene  glaciation 
(Denton  and  Hughes  1981). 
Ranges  of  the  stone  pines  (Mirov 
1967)  are  outlined  by  solid  or 
dashed  lines:  A  =  Pinus  cembra, 
B  =  P.  sibirica,  C  =  P.  pumila, 
D  =  P.  koraiensis,  E  =  P.  albicaulis. 
The  absence  of  stone  pines  in  the 
conifer  zone  of  northwestern 
Eurasia  and  parts  of  North  America 
could  be  due  to  glaciation. 


Ten  temperature  parameters  were  compared.  Midwin- 
ter frost  danger  was  described  by  the  long-term  minimum 
temperature  (the  absolute  low),  the  average  January 
minimum,  and  the  average  January  maximum.  Extreme 
fall  and  spring  frosts  are  represented  by  the  absolute  low 
in  the  first  and  last  months  of  winter  (that  is,  the  absolute 
lows  in  the  months  with  OC  average  air  temperature; 
Weaver  1994).  The  average  growing-season  temperature 
was  calculated  across  those  months  when  average  air 
temperatures  were  above  0  °C.  This  average  was  recalcu- 
lated using  "temperature  growth  support  units  =  Q"  to 
account  for  the  rise  in  rates  of  metabolic  processes  with 
rising  temperature;  these  points  lie  on  a  curve  defined 
by  0  °C  =  0,  1  °C  =  1,  11  °C  =  2,  21  °C  =  3,  31  °C  =  4,  and 
41  °C  =  5  (Weaver  1994).  Summer  highs  were  represented 
by  the  average  July  minimum,  average  July  maximum, 
and  long-term  high. 

Seven  precipitation  parameters  were  compared. 
October- June  precipitation  was  measured  to  determine 
whether  soil  was  moist  in  winter  and  at  the  opening  of 


the  growing  season.  Summer  precipitation  was  indexed 
by  July-September  precipitation  and  by  precipitation  in 
the  wettest  and  driest  month  in  that  period.  Since  plants 
require  water,  not  precipitation,  months  with  a  positive 
water  balance  were  estimated  with  the  Walter  index 
(Nielson  1986;  Walter  1973),  which  assumes  that  for  ev- 
ery 2  °C  rise  in  average  temperature  1  mm  of  precipita- 
tion will  be  evapotranspired.  Two  derivatives  were  stud- 
ied: drought  months  (duration  of  periods  with  a  negative 
water  balance)  and  drought  magnitude  (total  estimated 
deficit).  While  the  Walter  index  tends  to  underestimate 
drought  (Stephenson  1990),  it  may  be  adequate  in  these 
cool  climates.  Poikilohydric  (Larcher  1975)  organisms 
may  be  more  dependent  on  a  number  of  growing  hours  set 
by  the  number  of  rain  days  than  on  total  precipitation; 
rain  days  per  month  were  therefore  recorded. 

Four  growing-season  parameters  were  compared.  Length 
of  the  growing  season  was  indexed  first  as  the  number  of 
months  when  air  temperature  is  above  0  °C,  that  is,  the 
approximate  number  of  months  when  the  soil  is  thawed 


86 


(Weaver  1994).  Second,  drought  months  (none  according 
to  the  Walter  index)  were  subtracted  from  the  warm- 
season  index  to  create  a  better  warm-moist  season  index. 
Despite  contrary  observations  (Weaver  1994),  growth  is 
expected  to  be  indexed  by  the  integral  of  temperature  over 
growing  season.  Thus  two  indices  of  growing  season  pro- 
ductivity (W eaver  1994)  were  calculated:  [growing  season 
X  (average  growing  season  temperature  -5  °C)  (Chang 
1968)]  and,  considering  the  Q    effect,  [growing  season 
X  average  growing  season  Q]. 

WINTER  TEMPERATURES 

Average  winter  (January)  temperatures  in  northeastern 
Asia  are  the  world's  coldest  and  they  warm  southward 
(Fullard  and  Darby  1964).  Normal  daily  lows  in  regions 
occupied  by  the  trees  are  Pinus  pumila  (-30  °C),  P.  sibir- 
ica  (-27  °C),  P.  koraiensis  (-27  °C),  P.  albicaulis  (-14  °C), 
and  P.  cembra  (-8  °C),  respectively  (table  1).  Daily  highs 
in  winter  average  8  °C  higher  (table  1). 

Absolute  lows  experienced  by  Pinus  sibirica  (-55  °C) 
and  P.  pumila  (-52  °C)  rise  to  P.  koraiensis  (-42  °C), 
P.  albicaulis  (-34  °C),  and  P.  cembra  (-23  °C)  (table  1). 
Pinus  albicaulis  and  P.  cembra  may  be  protected  from  ex- 
treme low  temperatures  by  the  drainage  of  cold  air  from 
their  mountaintop  sites.  Becwar  and  Burke  1982  show 
that  timberline  conifers  of  the  Colorado  Rockies  do  not 
survive  temperatures  below  -40  °C;  thus  P.  sibirica  and 
P.  pumila  may  be  the  only  stone  pines  tolerant  of  -50  °C. 

Frosts  of  early  fall  and  late  spring  may  catch  trees  in 
partially  hardened  states.  Absolute  lows  in  the  first  and 
last  months  of  winter  were  -21  °C  for  P.  koraiensis,  -17  °C 
for  P.  sibirica,  and  -10  °C  for  the  remaining  trees.  Why 
are  absolute  lows  for  fall-winter-spring  frosts  lower  in  the 
P.  koriaensis  and  P.  sibirica  regions  than  in  the  P.  pumila 
regions  where  average  temperatures  are  lower  (table  1; 
Fullard  and  Darby  1964)? 


SUMMER  TEMPERATURES 

Continentality  allows  northeastern  Asia  to  warm  in 
summer  more  than  less  continental  parts  of  the  polar 
region.  Thus  absolute  maximum  temperatures  of  Pinus 
sibirica  (37  °C)  and  P.  pumila  (33  °C)  are  considerably 
higher  than  those  of  P.  cembra  (27  °C)  and  P.  albicaulis 
(29  °C)  (table  1).  Absolute  summer  highs  in  the  P.  korai- 
ensis forest  at  the  boreal-deciduous  forest  transition  are 
as  high  (36  °C)  as  in  interior  forests.  Average  July  highs 
in  P.  sibirica  and  P.  koraiensis  forests  are  like  those  in 
Rocky  Mountain  Douglas-fir  ( Pseudotsuga  menziesii )  for- 
ests, a  vegetation  zone  below  the  Rocky  Mountain  stone 
pine  zone  (Weaver  1994). 

Daily  growth  rates  are  determined  by  average  tempera- 
ture conditions  in  the  growing  season.  (1)  Temperature 
conditions  are  represented  most  simply  by  average  tem- 
perature; this  is  13  °C  for  Pinus  sibirica,  11  °C  for  P.  korai- 
ensis, and  8  to  9  °C  for  P.  pumila ,  P.  albicaulis,  and  P.  cem- 
bra. Like  July  highs,  13  °C  average  temperatures,  are 
similar  to  those  found  in  the  Douglas-fir/ponderosa  pine 
forests  of  the  Rocky  Mountains  (Weaver  1994).  (2)  Be- 
cause growth  rises  exponentially  with  increasing  tem- 
perature (Qjq;  Larcher  1975),  a  better  index  of  tempera- 
ture on  growth  may  be  a  cross-season  average  of  growth 
support  units  (Weaver  1994),  which  give  greater  weight 
to  high  than  low  temperatures.  With  this  index  climates 
cool  from  P.  sibirica  to  P.  albicaulis  to  P.  pumila-P.  korai- 
ensis to  P.  cembra  (table  1). 

PRECIPITATION 

Winter-spring  precipitation  is  high  in  Pinus  albicaulis 
(829  mm)  and  P.  cembra  (616  mm)  and  lower  (about 
250  mm)  in  northeastern  Asia  (table  2).  Since  the  lower 
amounts  will  saturate  most  moimtain  soils  (Weaver  1978), 
the  excess  is  expected  to  run  off,  to  have  little  effect  on 
water  supplies  in  forest  stands  during  the  growing  season. 


Table  1 — Temperatures 

^  in  stone  pine  communities  of  the  world. 

Climates  are  listed  in  order  of  increasing  winter  temperatures 

Species,  location,  number  of  stations 

2 

P .  pumila 

P.  sibirica 

P.  koraiensis 

P.  albicaulis 

P.  cembra 

Temperature 

N.  China 

Siberia 

Korea-China 

N.  America 

Euro-Alps 

data 

4+ 

4+ 

3 

4 

3 

Winter  Temperature 

Jan.  mean  min 

-30.0  ±  8.0 

-27.0  ±  2.0 

-27.0  ±  2.0 

-14.0  ±2.0 

-8.0  ±  0.0 

Jan.  mean  max 

-24.0  ±  9.0 

-19.0  ±2.0 

-11.0  ±2.0 

-5.0  ±  3.0 

-1.0  ±  1.0 

Abs  min 

-52.0  ±  4.0 

-55.0  ±  2.0 

-42.0  ±  1.0 

-34.0  ±  2.0 

-23.0  ±  1 .0 

Frost  spring 

-11.0  ±3.0 

-18.0  ±2.0 

-26.4  ±  0.2 

-1 1 .0  ±  1 .0 

-10.0  ±  1.0 

Frost  fall 

-11.0  ±3.0 

-16.0  ±2.0 

-27.6  ±  0.7 

-1 1 .0  ±  1 .0 

-10.0  ±  1.0 

Summer  Temperature 

July  mean  min 

8.0  ±  0.0 

12.0  ±  1.0 

14.0  ±1.0 

4.0  ±  1.0 

5.0  ±  1.0 

July  mean  max 

15.0  ±2.0 

21.0±  1.0 

26.0  ±  0.0 

18.0  ±  1.0 

14.0  ±1.0 

Abs  max 

33.0  ±1.0 

37.0  ±  1.0 

36.0  ±2.0 

29.0  ±  1.0 

27.0  ±  2.0 

9.0  ±  1.0 

13.0  ±1.0 

11.3±  1.2 

9.0  ±1.0 

8.0  ±  1.0 

1.9  ±0.2 

2.4  ±  0.1 

1.9  ±0.1 

2.1  ±0.2 

1.7  ±0.1 

'Temperature  data  (°C)  are  the  mean  ±  one  standard  error.  Absolute  temperatures  are  recorded  for  1 0  years  in  P.  albicaulis  and  P.  cembra; 
records  for  the  Asian  pines  are  unspecified  (presumed  longer).  T^,  and  Q^,  are  growing  season  averages  defined  in  the  text. 

^  Sample  size  is  four  for  P.  albicaulis  (except  average  max  and  min  for  January  and  July,  n  =  5),  three  for  P.  cembra,  and  for  both  P.  pumila  and 
P.  sibirica  eight,  except  for  January  max-min,  July  max-min,  and  spring-fall  frost  temperatures  where  n  =  4. 


87 


and,  thus,  to  have  httle  effect  on  production.  Large  snow- 
falls in  the  P.  albicaulis  and  P.  cembra  forests,  relative 
to  those  in  the  Asian  forests,  surely  result  in  greater 
snowpacks,  which  shelter  ground-level  plants  and  ani- 
mals less  frost  tolerant  than  the  trees. 

Summer  precipitation  is  lower  where  moisture  carrying 
air  masses  cross  mountains  (northeastern  Asia  and  the 
Rocky  Mountains)  than  where  they  do  not  (Alps  and  east- 
ern China).  Thus  summer  rainfall  increases  from  Pinus 
albicaulis  (102  mm)  to  P.  pumila  (187  mm),  P.  sibirica 
(187  mm),  P.  cembra  (323  mm),  and  P.  koraiensis 
(394  mm).  Precipitation  in  the  driest  month  parallels 
summer  rainfall  (table  2). 

Soil  water  may  be  adequate  in  spite  of  low  rainfall  if 
evapotranspiration  is  low.  If  we  assume  that  1  °C  degree 
evaporates  2  mm  per  month  (Nielson  1986,  1992;  Walter 
1973),  we  see  that  stone  pines  do  not  experience  drought 
in  any  month  (table  2).  In  addition,  some  buffering 
against  drought  occurs  because  snowmelt  water  stored  in 
the  soil  provides  a  supplement  to  summer  showers.  How- 
ever, even  at  boreal  temperatures  Walter's  index  may  un- 
derstate evaporation  (Stephenson  1990),  so  there  is  prob- 
ably less  summertime  restriction  of  growth  by  drought 
(stomate  closure)  in  Pinus  cembra  and  P.  koraiensis  than 
in  P.  albicaulis,  P.  sibirica,  and  P.  pumila  forests. 

Organisms  without  water  reserves — such  as  lichens, 
mosses,  and  invertebrates — are  more  sensitive  to  summer 
drought  and  may  distinguish  wetter  and  drier  forests. 
These  organisms  are  surely  inhibited  in  Pinus  pumila 
and  P.  albicaulis  forests  (one  to  eight  rain  days/month) 
relative  to  P.  cembra  and  P.  koraiensis  forests  (16  to  18 
rain  days). 

GROWING  SEASON 

Survival  depends  on  photos3aithetic  provisioning  for 
winter  respiration  and  the  outcome  of  competition  may 
depend  on  excesses  above  this  basic  provisioning.  One 
expects  production  to  be  correlated  with  the  number  of 
warm-moist  days,  with  the  warmth  of  those  days  (  Q^^, 
Larcher  1975),  and  thus  with  their  product. 


If  water  stress  never  occurs,  the  length  of  the  warm- 
moist  season  increases  from  Pinus  albicaulis  (4.5  mo) 
to  P.  pumila  (4.6  mo),  P.  sibirica  (5.5  mo),  P.  cembra 
(6.3  mo),  and  P.  koraiensis  (7.8  mo)  (table  3).  And  if  tem- 
perature conditions  were  identical  in  these  forests,  one 
would  therefore  expect  production  in  P.  cembra  and 
P.  koraiensis  forests  to  be  half  again  as  great  as  in 
P.  pumila  and  P.  albicaulis  forests.  Actual  differences 
could  be  even  larger  because  the  temperature-defined 
growing  season  may  overstate  stand  productivity  in  drier 
regions.  For  example,  while  subsoils  remain  moist  all 
summer  in  higher  (J.  Brown,  personal  commimication) 
and  lower  (Weaver  1974)  parts  of  the  P.  albicaulis  zone, 
drying  of  surface  soils  apparently  causes  tree  water  stress 
at  lower  (B.  Keane,  personal  communication),  but  not 
higher  (J.  Brown,  personal  communication),  sites.  Similar 
droughty  periods  probably  occur  in  warmer  P.  pumila 
and  P.  sibirica  forests. 

Production  is  expected  to  rise  exponentially  with  rising 
temperature,  and  thus  productivity  might  be  better  corre- 
lated with  the  product  of  season  length  and  temperature 
"growth  supporting  units."  While  this  production  hy- 
pothesis was  rejected  in  a  cross-vegetation  zone  analysis 
(Weaver  1994),  the  genetic  similarity  of  stone  pines  might 
allow  it  to  operate  here.  If  so,  and  if  water  deficits  in 
Siberia  do  not  shorten  the  growing  season,  P.  koraiensis 
and  P.  sibirica  will  be  promoted  to  the  most  productive 
stone  pines  (table  3). 

CONCLUSIONS 

From  a  tree's  point  of  view,  average  conditions  in  the 
stone  pine  zones  are  similar:  water  stress  is  slight  or 
nonexistent  and  growing  season  temperatures  average 
9  to  13  °C.  Seasonal  extreme  conditions  may,  however, 
prevent  reciprocal  transplantation.  Winter  or  fall-spring 
frosts  in  the  Pinus  sibirica/P.  pumila/P.  koraiensis  region 
may  exclude  the  other  pines.  And  weak  droughts  in  the 
P.  albicaulis/P.  pumila/P.  sibirica  regions  might  exclude 
P.  cembra  or  P.  koraiensis. 


Table  2 — Precipitation^  in  stone  pine  forests  of  ttie  world.  Climates  are  listed  in  order  of  increasing  summer  precipitation 


Species,  location,  number  of  stations  ^ 

P .  albicaulis 

P.  pumila 

P.  s  ib  irica 

P.  cembra 

P.  koraiensis 

Precipitation 

N.  Annerica 

N.  China 

Siberia 

Euro-Alps 

Korea-China 

data 

4 

4+ 

4+ 

3 

3 

Total 

931  ±  229 

407  ±137 

432  ±  21 

939  ±  9 

636  ±  68 

Oct.-June 

829  ±  234 

264 ±  76 

245  ±  22 

616  ±43 

242  ±  62 

July-Sept. 

102  ±  14 

143  ±  43 

187  ±11 

323  ±  36 

394  ±  17 

Wettest  summer  month 

116±  16 

165 ±  25 

181  ±  8 

214±  15 

346 

Driest  summer  month 

4±  4 

4±  2 

8±  2 

45  ±16 

32 

Summer  drought  months 

0±  0 

0±  0 

0±  0 

0±  0 

0±  0 

Summer  water  deficit 

0±  0 

0±  0 

0±  0 

0±  0 

0±  0 

Summer  rain  days, 

average  number 

8±  1 

1  ±  1 

14±  1 

16±  1 

18 

'Precipitation  (mm)  data  are  total  (sum  of  all  months),  winter,  summer  (July,  August,  plus  September),  driest  summer  month  (July-September)  recorded, 
wettest  summer  month  (July-September)  recorded,  and  average  number  of  showers  in  June-September.  High  variances  in  total  precipitation  for  P. 
albicaulis  and  P.  pumila  are  reduced  to  705  ±  51  mm  and  274  ±  37  mm  by  omission  of  the  Crater  Lake  and  Petropavlovsk  stations,  respectively. 

^Sample  size  is  four  for  P.  albicaulis,  three  for  P.  cembra,  three  for  P.  Koraiensis,  and  eight  for  both  P.  pumila  and  P.  sibirica,  except  for  summer  drought 
months  and  deficit,  where  it  is  only  four. 


88 


Table  3 — Stone  pine  productivity  correlates:  growing  season  ,^  growing  season  temperature,^  and  their  products.  Climates  are  listed  in 
descending  order  by  one  estimate  of  productivity,  warm  moist  season 

 Species,  location,  number  of  stations  

P.  koralensis  P .  cembra  P.  siblrica  P.  pumlla  P.  albicaulis 

Korea-China  Euro-Alps  Siberia  N.  China  N.  America 


4 

4 

4 

Warm  season 

7.8  ±0.1 

6.3  ±  0.2 

5.5  ±  0.2 

4.6  ±  0.4 

4.5  ±  0.3 

Summer  drought  months 

0.0  ±0.0 

0.0  ±  0.0 

0.0  ±0.0 

0.0  ±0.0 

0.0  ±0.0 

Warm-moist  season 

7.8  ±0.1 

6.3  ±  0.2 

5.5  ±  0.2 

4.6  ±  0.5 

4.5  ±  0.3 

k 

GS  X  Tg -5  °C 
GSxQg, 

11.3±1.2 

8.0  ±1.0 

13.0  ±  1.0 

9.0  ±  1.0 

9.0  ±  1.0 

1.9  ±0.1 

1.7  ±0.1 

2.4  ±0.1 

1.9  ±0.2 

2.1  ±0.2 

56.5 

18.9 

44.0 

18.4 

18.0 

15.6 

10.7 

13.2 

8.7 

9.5 

'Warm-season  and  warm-moist  season  are  two  indices  of  growing  season.  Warm-season  months  occur  after  average  monthly  air  temperature  hses 
above  0  °C  and  before  it  falls  below  0  °C.  The  warm-moist  season  Is  the  warm  season  minus  any  months  in  which  T/2  is  greater  than  P(mm),  Walter  1973; 
"Walter  drought"  does  not  occur  in  stone  pine  regions. 

H'wo  indices  of  growing-season  temperature  are  provided.  T^,  is  the  average  temperature  in  growing  season  months.  T  -5C  is  used  as  one  index  of 
growth  support  on  the  assumption  that  growth  does  not  occur  below  5  °C  (Chang  1 968).  Q  is  the  average  of  temperatures  weighted  tor  their  growth- 
supporting  capacity:  0  °C  =  0,  1  °C  =  1 ,  1 1  °C  =  2,  21  °C  =  4  {Weaver  1993). 


Organisms  filling  other  niches  might  find  the  climates 
dissimilar  (Weaver  1990).  Poikilohydric  organisms  and 
himaans  would  react  differently  to  the  relatively  smnmer- 
rainy  climates  of  the  P.  cembra  and  P.  koraiensis  regions 
than  to  the  other  stone  pine  climates.  And  organisms 
wintering  imder  snow — small  mammals,  insects,  plants 
of  small  stature — are  more  protected  from  the  extremes 
of  winter  cold  in  the  P.  cembra  and  P.  albicaulis  regions 
than  in  the  other  climates. 

ACKNOWLEDGMENTS 

I  thank  (1)  W.  Tranqmllini  (University  of  Innsbruck, 
Austria)  for  providing  data  from  his  P.  cembra  weather 
stations,  (2)  J.  Luo  (Beijing  Forestry  University,  China) 
and  Z.  Shidong  (Inst.  Appl.  Ecol.,  Shenyang,  China)  for 
providing  data  from  weather  stations  in  P.  koraiensis  cli- 
mates, (3)  H.  Lieth  (University  of  Osabuck,  Germany), 
P.  Gorchakovsky  (Acad.  Sci.  Inst.  Ecol.,  Ekaterinburg, 
Russia)  and  P.  Khomentovsky  (Kamchatka  Inst.  Ecol., 
Petropavlovsk,  Russia)  for  help  in  selecting  P.  sibirica 
and  P.  pumila  weather  stations,  and  (4)  nineteen  North 
American  ecologists  already  acknowledged  (Weaver  1990) 
for  help  in  selecting  North  American  P.  albicaulis  stations. 
I  am  also  grateful  for  support  to  the  project  provided 
through  Intermountain  Research  Station  contract  #INT- 
92720-RJVA  (W.  Schmidt). 

REFERENCES 

Becwar,  M.;  Burke,  M.  1982.  Winter  hardiness  limitations 

and  physiography  of  woody  timberline  flora.  In:  Li,  P.; 

Sakai,  A.  Plant  cold  hardiness  and  freezing  stress.  New 

York:  Academic  Press:  307-324. 
Chang,  J.  1968.  Climate  and  agriculture.  Chicago:  Aldine. 

304  p. 

Fullard,  H.;  Darby,  H.  1964.  The  imiversity  atlas.  London: 
George  Philip  and  Son.  176  p. 

Lanner,  R.  1990.  Biology,  taxonomy,  evolution,  and  ge- 
ography of  stone  pines  of  the  world.  In:  Schmidt,  W.; 
McDonald,  K.,  comps.  Proceedings — symposium  on 


whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  14-24. 

Larcher,  W.  1975.  Physiological  plant  ecology.  New  York: 
Springer.  252  p. 

Mirov,  N.  1967.  The  genus  Pinus.  New  York:  Ronald 
Press.  602  p. 

Muller,  M.  1982.  Selected  climatic  data  for  a  global  set 
of  standard  stations  for  vegetation  science.  The  Hague, 
Netherlands:  Junk.  306  p. 

Neilson,  R.  1986.  High  resolution  climatic  analysis  and 
southwest  biogeography.  Science.  232:  27-33. 

Nielsen,  R.  1992.  Toward  a  rule  based  biome  model.  Land- 
scape Ecology.  7:  27-43. 

Stephenson,  N.  1990.  Climatic  control  of  vegetation  distri- 
bution: the  role  of  water  balance.  American  Naturalist. 
135:  649-670. 

Walter,  H.  1973.  Vegetation  of  the  earth  in  relation  to  cli- 
mate and  ecophysiolgical  conditions.  New  York: 
Springer.  237  p. 

Weaver,  T.  1974.  Root  distribution  and  soil  water  regimes 
in  nine  habitat  types  of  the  northern  Rocky  Mountains. 
In:  Marshall,  J.,  ed.  The  belowgroimd  ecosystem.  Sci. 
Ser.  26.  Fort  ColHns,  CO:  Colorado  State  University, 
Range  Science  Department.  351  p. 

Weaver,  T.  1979.  Changes  in  soils  along  a  vegetational 
(altitudinal)  gradient  of  the  northern  Rocky  Mountains. 
In:  Youngberg,  C,  ed.  Proceedings  of  the  Fifth  North 
American  forest  soils  conference.  Madison,  WI:  Soil  Sci- 
ence Society  America.  14-29. 

Weaver,  T.  1990.  Climates  of  subalpine  pine  woodlands. 
In:  Schmidt,  W.;  McDonald,  K.,  comps.  Proceedings — 
s)rmposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-moimtain  resource.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
72-79. 

Weaver,  T.  1994.  Vegetation  distribution  and  production 
in  Rocky  Mountain  climates — wdth  emphasis  on 
whitebark  pine  [these  proceedings]. 


89 


Growth  Characteristics 


International  Workshop 
St.  Moritz  1 992 


90 


GROWTH  OF  SWISS  STONE  PINES 
THAT  ORIGINATED  FROM  AND  WERE 
PLANTED  AT  SEVERAL  ALTITUDES  IN 
THE  AUSTRIAN  ALPS 

Kurt  Holzer 


Abstract — For  the  purpose  of  gene  conservation,  a  plus  tree 
collection  of  Swiss  stone  pine  {Pinus  cembra)  was  conducted. 
Scions  of  185  trees  and  82  open-pollinated  seed  samples  were 
collected  within  the  natural  range  of  Swiss  stone  pine  in  Austria, 
In  1987,  clonal  height  was  measured  at  an  age  between  25  and 
30  years.  Open-pollinated  families  were  planted  at  three  differ- 
ent sites.  Height  was  measured  after  18  years.  Growth  of  graft- 
ings and  open-pollinated  families  helped  identify  the  altitude  of 
their  origin.  When  planted  below  timberline,  growth  decreased 
with  increasing  altitude. 


In  1957,  the  Department  of  Forest  Tree  Breeding  and 
Genetics,  Federal  Forestry  Research  Institute,  Austria, 
commenced  a  selection  program  of  Swiss  stone  pine  (Pi- 
nus cembra).  The  objectives  of  this  program  were  to 
preserve  genetic  resovirces  of  superior  trees  in  a  clonal 
orchard  and  to  assess  the  growth  capacity  of  the  clones 
by  field  performance  of  open-pollinated  families. 

Within  the  natural  range  of  stone  pine,  timberline  di- 
vides the  distribution  of  stone  pine,  due  to  environmental 
conditions,  into  two  silvicultural  zones:  (1)  production  for- 
ests and  (2)  "Kampfzone."  Hence,  those  aspects  had  to  be 
considered  for  the  program  (Holzer  1963, 1976). 

MATERIAL  AND  METHODS 

Covering  the  majority  of  natural  stands  in  Austria, 
185  superior  trees  were  selected  between  1958  and  1965 
(Holzer  1961,  1969).  Within  production  forests  scions  of 
124  and  above  timberline  ("Kampfzone")  scions  of  61  indi- 
viduals were  collected.  The  altitudinal  range  was  approxi- 
mately 250  m  below  and  230  m  above  timberline  (be- 
tween 1,650  m  and  2,200  m  above  sea  level).  In  addition, 
open-pollinated  seeds  were  harvested  from  82  trees. 

Graftings  were  planted  in  a  conservation  orchard  in 
Purkersdorf-Stadlhutte  near  Vienna  at  400  m  above  sea 
level.  Progenies  were  raised  in  the  nursery  of  the  insti- 
tute (Mariabrvmn  Vienna,  220  m  above  sea  level)  for 
6  to  8  years  and  then  planted  at  three  different  sites: 
(1)  Purkersdorf-Stadlhutte;  (2)  GroBe  Zirbenwiese  (1,650  m 
above  sea  level);  Seetaler  Alpen,  a  swampy  meadow 
within  the  natiiral  range  of  stone  pine,  approximately 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Kurt  Holzer  is  Professor  of  Forest  Genetics,  Institut  fur  Waldbau, 
Universitat  fur  Bodenkultur,  A-1190  Vienna,  Austria. 


250  m  below  timberline;  and  (3)  Melcheben,  Packalpe 
(1,700  m  above  sea  level),  close  to  timberline  of  Norway 
spruce  (Picea  abies).  At  sites  2  and  3,  only  32  families 
were  planted  because  of  limited  nvmabers  of  plants. 

In  1987,  height  was  measvired  at  a  clonal  age  between 
25  and  30  years.  Heights  of  progenies  were  assessed  at 
an  age  of  8  and  18  years,  respectively.  Exclusively  de- 
scriptive statistics  were  employed. 

Some  results  have  been  previously  published  with 
respect  to  graftings  (Holzer  1989)  and  progenies  (Holzer 
1978).  Recently,  growth  of  vegetative  and  generative 
plant  material  has  been  measured  (Feuersinger  1992). 

Since  altitude  of  timberline  is  irregular  within  the 
Alps,  comparisons  are  based  on  relative  distances  to  local 
timberline. 

RESULTS  AND  DISCUSSION 

Results  are  summarized  in  figures  1-3  within  the  main 
distribution  (about  150  m  below  to  100  m  above  timber- 
line);  mean  height  of  30-year-old  graftings  decreased  with 
increasing  elevation  of  origin  fi-om  3.8  m  at  150  m  below 
to  2.7  m  at  100  m  above  timberline.  Variation  was  high 
and  yielded  to  50  to  70  percent  of  the  clonal  mean.  In 
general,  the  majority  of  graftings  originating  below  tim- 
berline had  superior  growth.  Presimiably,  inherent 
growth  capacity  shows  a  cline  within  the  main  distribu- 
tion zone  of  stone  pine  (fig.  1). 

This  growth  pattern  was  also  found  in  generative  plant 
material.  Progenies  originating  fi'om  100  m  below  timber- 
line  showed  pronoimced  reduced  growth.  However,  when 
growth  was  only  evaluated  within  a  single  seed  zone,  it 
was  always  higher  below  the  timberline  than  above  (fig.  2). 
In  the  nursery,  height  of  families  (n  =  50)  originating  fi'om 
production  forests  averaged  46.5  cm.  Families  in  =  16)  fi'om 
the  "Kampfzone"  had  a  slightly  reduced  growth  of  40.0  cm 
at  an  age  of  8  years. 

Transplants  of  families  18  years  old  showed  different 
results.  Clinal  growth  pattern  is  still  distinguishable  at 
site  1  (400  m  above  sea  level)  and  site  2  (250  m  below  tim- 
berline). However,  at  site  3  (timberline)  height  growth  re- 
actions were  different.  Here,  families  with  small  height 
growth  potential  at  400  m  above  sea  level  were  superior. 
Figure  3  shows  growth  of  families  originating  fi'om  two 
different  seed  zones.  In  seed  zone  "Defi-eggen"  growth 
of  18  families,  400  m  and  1,700  m  above  sea  level,  was 
negatively  correlated  (r  =  -0.752***).  Fovuteen  families 
originating  from  seed  zone  "Prankerhohe"  showed  a  con- 
sistent pattern.  However,  correlation  (r  =  -0.205)  was  not 
significant. 


91 


200 -X 

150- 

100- 

? 

£  50- 

O) 

O  above 

f timber  line 
below 

50  - 
100- 

150- 
200- 
250- 


o 

3 


O  X  X  J     X  X  S 

I 


fix  X 


-|^T<  5-5-xW^X-X-^-> 


XXX  X.J 

X  X  Sx  X  S  X  X  X  >j6^  I  i  ^ 

N 

XX  Jgx      X^xg)^  X 

I  \ 

X  xx8xi>x'*8    X  X 

I  X 

/ 

X       X  I     XX  X 

X  S  I   X  P         XX  X 

/ 

l/ 

I  I  I 
3  4 

Height  of  Graftings  (m) 


Figure  1 — Total  height  growth  of  30-year-olcl 
graftings  In  the  plantation  at  400  m  above  sea 
level  with  the  altitude  of  their  origin  (distance 
from  the  local  timberline).  Dotted  curve  denotes 
the  mean  values. 


130 


120 


110  - 


100 


90  - 


1 ,700  m  asl 

X 


o  o 


X  \ 

x"^  X 

X  \ 

\ 


X 

o 


\ 


400  m  asl 


100 


120  140 

Height  (cm) 


160 


Figure  3 — The  comparison  of  the  mean  of  the 
1 8-year-old  families  on  two  different  planting 
sites:  400  m  and  1 ,700  m  above  sea  level,  re- 
spectively (provenance  1/6/100,  Prankerhohe, 
marked  with  o,  provenance  1/8/100,  Defreggen, 
marked  with  x;  see  text  also).  Curve  denotes 
correlation  (r  =  -0.752)  of  the  Defreggen  zone. 


This  observation  is  supported  by  a  sowing  trial  200  m 
above  local  timberline  (2,100  m  above  sea  level).  Pro- 
genies originating  from  sites  above  timberline  performed 
well;  progenies  from  trees  in  production  forests  performed 
poorly  (Holzer  1975). 


What  conclusions  regarding  growth  of  stone  pine  might 
be  drawn  from  these  experiments  so  far? 

1.  Height  growth  of  stone  pine  shows  clinal  variation. 

2.  When  plants  are  grown  in  production  forests,  growth 
is  positively  correlated  to  elevation. 

3.  Above  timberline  this  correlation  is  negative. 


200- 
150- 
100- 


•^50-  e 
c 

S above 
timber  line 
o  below 


•E  50- 
< 

100- 


4Q  0 

+ 


-X-XQW-^ 


1s 


o  +        ++  oma  _pco>ooo  0    o    o  o 


+  +  +  Ot  + 


150- 


200- 


■H-X  |X         X"xp        X  XX 


30 


-H-O-l-  +■ 

I 

40 


50  60 

Height  of  Progenies  (cm) 


70 


Figure  2 — Total  height  of  the  8-year-old  prog- 
enies (mean  of  families)  of  the  harvested 
clones  with  the  altitude  of  their  origin  (distance 
from  the  local  timberline).  Dotted  curve  de- 
notes the  mean  values. 


REFERENCES 

Feuersinger,  P.  1992.  Das  Wachstum  von  Zirbe  in 
Abhangigkeit  vom  Ursprungsort.  Diplomarbeit,  Univ.  f. 
Bodenkultur,  Wien.  140  p. 

Holzer,  K.  1961.  Der  derzeitige  Stand  der  Auswahl  von 
Schwarzerlen  und  von  Zirbe  im  Rahmen  der 
Forstpflanzenziichtung.  Informationsdienst  50,  Allg. 
Forstztg.  72. 

Holzer,  K.  1963.  Pinus  cembra  L.  as  a  pioneer  at  timber- 
line  in  the  European  Alps.  Proc.  World.  Cons.  Forest 
Genet,  and  Tree  Imprv.,  Vol.  1  (FAO/FORGEN  1963, 
3/13,  FAO,  Rome). 

Holzer,  K.  1969.  Erste  Ergebnisse  von  Zirbenauswahl- 
baumen  {Pinus  cembra  L.).  Centrbl.f.d.  ges.  Forstwes. 
86:  149-160. 

Holzer,  K.  1975.  Genetics  of  Pinus  cembra  L.  Annales 

Forestales,  Zagreb.  6(5):  139-158. 
Holzer,  K.  1976.  Breeding  haploxylon  pines  for  subalpine 

regions.  XVI  lUFRO-Congr.  Oslo.  Div.  II:  216-227. 
Holzer,  K.  1978.  Versuchsanlage  GroBe  Zirbenwiese.  Exk. 

Fiihrer  Arbeitsgem.  f.  Hochlagenaufforstung.  5  p. 

[Mimeo.] 

Holzer,  K.  1989.  Drei  Jahrzehnte  Erfahrungen  mit 
Zirbenpfropfimgen.  Centrbl.  f.d.  ges.  Forstwes.  106: 
79-88. 


92 


SEASONAL  PATTERNS  OF  GROWTH 
AND  PHOTOSYNTHETIC  ACTIVITY  OF 
PINUS  PUMILA  GROWING  ON  THE  KISO 
MOUNTAIN  RANGE,  CENTRAL  JAPAN 

Takuya  Kajimoto 


Abstract — Seasonal  patterns  of  growth  and  photosynthesis  in 
Pinus  pumila  needles  were  investigated  in  relation  to  carbon 
uptake.  Current  needles  elongated  from  July  to  late  August. 
Photosynthetic  activities  of  2-year-old  needles,  both  in  sun-  and 
shade-needles,  reached  seasonal  maxima  in  August,  while  those 
of  current  needles  did  so  in  September.  The  results  suggest  that 
annual  net  carbon  gain  of  P.  pumila  depended  mainly  on  net  pho- 
tosynthesis of  old  needles  (1  to  4  year  olds). 


Growth  and  production  rates  of  trees  within  the  timber- 
Une  ecotone  are  generally  restricted  mainly  due  to  the 
short  growing  season  (Tranquillini  1979).  However,  most 
of  the  subalpine  species  show  physiological  adaptations  to 
conditions  associated  with  low  temperature,  lower  photo- 
synthetic  optimun  temperature  (Larcher  1975),  and 
higher  frost  and  freezing  resistance  (Sakai  and  Larcher 
1987).  It  has  also  been  suggested  that  within  krummholz 
mats  consisting  of  Picea  engelmannii  and  Abies  lasiocarpa 
photosynthesis  is  enhanced  by  more  favorable  canopy- 
temperature  conditions  (Hadley  and  Smith  1987).  Infor- 
mation on  such  physiological  and  structural  functions  is 
indispensable  for  understanding  how  alpine  trees  survive 
by  achieving  positive  carbon  balance  during  the  short 
siunmers. 

In  Japan,  Pinus  pumila  Kegel,  occurs  in  high-moimtain 
areas  from  central  Honshu  (the  main  island)  to  the  north- 
em  island  of  Hokkaido.  This  pine  generally  exhibits 
dwarfed  growth  forms,  and  regenerates  by  layering  at 
the  mature  growth  stage  (Kajimoto  1992;  Okitsu  and  Ito 
1984).  The  physiognomy  of  P.  pumila  is  similar  to  Pinus 
mugo  at  the  upper  timberline  in  the  European  Alps  or  in 
the  Carpathian  Mountains  (Holtmeier  1973,  1981;  Wardle 
1977). 

The  foliage  biomass  of  P.  pumila  rsinges  from  15  to  24 
tons  d.w./ha  despite  the  low  tree  heights.  Pinus  pumila 
stands  develop  dense  canopies  with  leaf  area  density  of 
about  5  mVm^,  which  is  considerably  more  than  in  other 
conifer  forests  (Kajimoto  1989a).  This  canopy  structure 
may  result  in  large  annual  net  production  (Okitsu  and  Ito 
1989;  Shidei  1963).  However,  seasonal  changes  in  growth 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Takuya  Kajimoto  is  a  special  Researcher  of  Science  and  Technology, 
Laboratory  of  Plant  Production,  Forestry  and  Forest  Products  Research 
Institute,  Tsukuba,  Ibaraki  305,  Japan. 


and  photosynthesis  of  P.  pumila  needles  are  still  un- 
known in  relation  to  their  carbon  uptake. 

In  this  paper,  a  process  of  photosynthetic  production 
in  P.  pumila  during  the  growing  season  is  discussed 
based  on  observations  of  environmental  factors  and 
seasonal  patterns  of  growth  and  photosynthesis  in  the 
needles. 

STUDY  AREA 

The  study  area  is  located  on  the  Kiso  Mountain  Range 
in  central  Japan.  The  range  trends  from  south  to  north. 
The  highest  peak  is  Mount  Kisokoma  (2,965  m  a.s.l.). 
In  this  mountain  range,  P.  pumila  trees  naturally  occur 
above  2,500  m.  The  study  was  carried  out  in  the 
Shinshu  University  Experimental  Forest  (35°48'  N., 
137°50'  E.)  in  the  northern  part  of  the  Kiso  Mountain 
Range.  The  investigated  pine  stand  is  located  on  the 
northeast  slope  (15°  inclination)  at  an  altitude  of 
2,600  m  a.s.l.  Dense  canopy  layer  is  developed  between 
100  and  200  cm  above  the  ground  height  (Kajimoto 
1989a). 

FIELD  OBSERVATIONS 

Four  siin-exposed  terminal  shoots  of  P.  pumila  were 
selected.  Lengths  of  current  shoots,  current  needles, 
and  winter  buds  were  measured  seven  times  between 
late  May  and  October  1988.  Current  needle  length  of 
each  shoot  was  determined  as  the  average  of  10  fascicles 
selected  at  equidistant  intervals  along  the  shoot  axis. 

At  the  plot,  solar  irradiance  (400-1,100  nm  in  wave- 
length) above  the  canopy  siuface  was  measiu-ed  with  a 
pyranometer  sensor  (LICOR,  LI-200SB)  at  1-hour  inter- 
vals from  June  to  October  1988.  Photos3nithetic  photon 
fliix  density  (PFD)  (400-700  nm)  above  the  canopy  sur- 
face and  air  temperature  inside  the  stand  were  also 
recorded  at  1-hour  intervals  between  mid-July  and 
October  1988  using  a  quantum  sensor  (Koito,  IKS-25) 
and  a  resistance  thermometer  (Koito,  OPT-150). 

MEASUREMENT  OF 
PHOTOSYNTHESIS 

Rates  of  COg  exchange  were  determined  on  detached 
shoots  that  were  collected  50  hours  prior  to  the  meas- 
urement. The  shoots  were  carefully  transported  to  the 
laboratory  to  minimize  error  associated  with  excision 
and  transportation  (Kajimoto  1990).  An  open  system 


93 


with  an  infrared  gas  analyzer  (Hartmann  and  Braun, 
URA-2)  was  used.  The  air  was  fed  into  an  assimilation 
chamber  (9  by  25  by  1  cm^)  at  a  rate  of  O.Sl/min.  COg  con- 
centration of  the  input  air  was  within  the  range  of  360- 
400  ppm.  The  chamber  was  placed  in  a  thermoregulated 
waterbath  and  illimiinated  with  incandescent  lamps. 
PFD  at  the  chamber  surface  was  measured  with  a  quan- 
tum sensor  (LICOR,  LI-190SB).  Details  of  the  system 
were  described  by  Kajimoto  (1990). 

Two  terminal  shoots  (about  10  cm  in  length)  of  P. 
pumila  were  sampled  from  the  sun-exposed  canopy  sur- 
face part  (1.8-2.0  m)  and  in  the  lower  shaded  canopy  lay- 
er (1.0-1.2  m)  seven  times  between  late  May  and  October 
1988.  At  each  sampling  time,  about  26  fascicles  of  ciirrent 
needles  (flushed  in  1988)  and  2-year-old  needles  (flushed 
in  1986)  were  separately  removed  from  each  sun-  and 
shade-shoot.  The  needles  were  placed  in  the  assimilation 
chamber  keeping  the  cut  end  of  each  fascicle  immersed  in 
water  using  a  small  vinyl  tube.  Net  photosynthetic  rates 
at  different  PFD  levels  (20-1,300  ^imol/mVs)  and  dark  res- 
piration rates  of  these  two  needle  groups  were  measured. 
Air  temperature  inside  the  chamber  was  maintained  at 
10  °C  throughout  the  measurement;  this  temperature 
value  was  within  the  range  of  optimum  air  temperatiire 
(10-15  °C)  in  net  photosynthesis  for  P.  pumila  needles 
(Kajimoto  1990). 


25 


20  - 


15 


o 


<  0 

-5 

-10 
25 


max. 


c 
o 


20  - 


1^ 


10 


£^  5 
o 

^  0 


M 


J  A 
1988 


0  N 


Figure  1 — Seasonal  changes  in  daily  solar  irradiance 
(400-1 ,100  nm  in  wavelength)  and  air  temperatures 
during  1988  measured  in  a  P.  pumila  stand  (2,600  m 
a.s.l.),  Kiso  Mountain  Range.  Each  point  stands  for 
the  mean  value  at  1  -week  intervals. 


£ 

J3 


V) 

■*-' 
C 
Ct 

V. 

O 

o 

U) 


C 

o 

Q> 

"8 

c 


—I  O 


Figure  2 — Seasonal  growth  patterns  of  current 
shoots,  current  needles,  and  winter  buds  ob- 
served for  sun-exposed  P.  pumila  shoots.  Each 
point  shows  the  mean  for  n  =  4  shoots,  o  = 
shoot  length;  •  =  needle  length;  □  =  bud 
length. 


Needle  area  of  each  sample  was  measured  as  its  pro- 
jected area  with  an  area  meter  (Hayashi,  AAC-100),  and 
its  dry  weight  was  measured  after  ovendrying  at  85  °C. 
Specific  leaf  area  (SLA;  cm7g)  was  determined  as  the  ra- 
tio of  needle  area  to  dry  weight.  Net  assimilation  rates 
obtained  in  all  measurements  were  converted  to  the  val- 
ues at  a  COg  concentration  of  380  ppm. 

RESULTS 

Climatic  Conditions 

Daily  solar  radiation  at  the  P.  pumila  stand  was  higher 
in  June  and  gradually  decreased  until  October  in  1988 
(fig.  1).  Mean  air  temperature  was  about  12  °C  during 
August  and  fell  below  0  °C  after  mid-October.  Monthly 
precipitation  was  212  mm  in  July  and  167  mm  in  August 
1988,  according  to  the  data  obtained  at  Senjoujiki  Station 
(2,623  m  a.s.l.)  in  the  central  part  of  Kiso  Mountain  Range 
(Nagano  Meteorological  Observatory  1988). 

Growth  Patterns  of  Shoots  and 
Needles 

The  current  shoots  elongated  rapidly  from  June  to  July 
(fig.  2).  The  current  needles  began  to  flush  in  early  July 
and  elongated  rapidly  during  August.  In  late  May,  the 
winter  buds,  which  would  flush  in  the  next  season,  were 
visible.  The  length  of  winter  bud  increased  gradually  un- 
til late  October. 

Specific  leaf  area  (SLA)  of  the  current  needles,  both 
sun-  and  shade-needles,  was  quite  large  in  early  August 
when  the  needles  began  to  grow  larger,  and  sharply  de- 
creased by  mid-September  (fig.  3).  SLA  of  2-year-old 


94 


Cur -Shade 


Figure  3 — Seasonal  changes  in  specific  leaf  area 
(SLA)  of  the  P.  pumila  needles  used  for  photosyn- 
thesis measurements.  □  =  current,  sun-needles; 
■  =  current,  shade-needles;  O  =  2-year-old,  sun- 
needles;*  =  2-year-old,  shade-needles. 


needles  became  slightly  larger  in  August  than  in  other 
months.  The  shade-needles,  both  current  and  2-year-old, 
showed  larger  SLA  values  throughout  the  season  than 
sun-needles. 


Cur.- Sun 


/ 


Cur-Shod  e 
2  yr-Sun 


2yr-Shade 


1000  1200 


Photosynthetic  photon  flux  density 
(pmol  nn-2s~i ) 

Figure  A — Examples  of  the  relationships  between 
photosynthetic  photon  flux  density  (PFD)  and  net 
photosynthetic  rate  (PJ  at  10  °C  of  P.  pumila 
needles  on  September  15, 1988.  □  =  current,  sun- 
needles;  ■  =  current,  shade-needles;  0=  2-year-old, 
sun-needles;  •  =  2-year-old,  shade-needles. 


PFD 


Light  Response  of  Net  Photosynthesis 

Light  response  of  photosynthesis  was  examined  based 
on  the  relationships  between  photon  flux  density  (PFD) 
and  net  photosynthetic  rates  (PJ  at  10  °C  (fig.  4).  Light 
saturation  points  of    were  different  between  sun-  and 
shade-needles,  and  also  between  needles  of  different  age. 
However,  all  sampled  needles  were  completely  light  satu- 
rated in     at  PFD  value  below  1,000  |imol/m7s  and  10  °C. 
A  maximum  net  photos5nnthetic  rate  (P^^)  was  defined  as 
P^  value  at  1,000  |imol/mVs  and  10  °C,  and  an  approxi- 
mate light-saturation  point  in  net  photosynthesis  (PFDg^) 
was  also  determined  as  PFD  value  that  gave  90  percent 
ofP  . 

max 

Table  1  shows  that  PFDgg  of  2-year-old  needles  ranged 
from  180  to  500  |imol/mVs  in  sun-needles  and  180  to 
440  |imol/m7s  in  shade-needles.  The  values  were  rela- 
tively lower  than  those  of  the  current  needles,  where 
PFDgo  was  440-580  [imol/mVs  in  sun-needles  and  300-500 
\imo]/mVs  in  shade-needles.  Both  current  and  2-year-old 
needles  had  highest  PFDgg  values  in  August  compared  to 
the  other  months. 

Seasonal  Change  in  Photosynthetic 
Activity 

Figure  5  shows  that  P^^  of  current  needles  were  larger 
in  sun-needles  than  in  shade-needles  for  each  period.  For 
2-year-old  needles,  there  was  a  remarkable  difference  in 

between  sun-  and  shade-needles  in  July  and  August. 
The  values  of  sun-  and  shade-needles  both  became  higher 
in  August  than  in  other  months.  P    values  of  current 

^  max 

needles  increased  between  late  August  and  mid-September. 

Dark  respiration  rates  at  10  °C  (R)  of  current  needles 
were  larger  in  sun-needles  than  in  shade-needles,  while 
there  were  no  differences  in  2-year-old  sun-  and  shade- 
needles  (fig.  5).  Current  needles  showed  larger  R  values 
than  2-year-old  needles  throughout  the  season. 


Table  1 — Approximate  light-saturation  points  (PFDgg)  in  the  net 
photosynthetic  rates  (at  1 0  °C)  of  P.  pumila  needles  in 
1988 

Light-saturation  points  (PFD^Y 


Date 

Current  needles 

2-year-old-needles 

Sun 

Shade 

Sun 

Shade 

May  27 

420 

440 

July  2 

340 

340 

July  20 

300 

300 

August  4 

500 

500 

500 

380 

August  24 

580 

460 

480 

380 

September  1 5 

440 

360 

180 

180 

October  20 

500 

300 

240 

220 

'PFDgg  was  defined  as  the  value  of  photosynthetic  photon  flux  density 
which  gave  90  percent  of  a  maximum  net  photosynthetic  rate  {P^^J-  Each 
PFDg^  value  was  determined  using  the  light-photosynthetic  curve,  as  shown  in 
figure  4. 


95 


«»  12 

«A 

*  ft, 

c  o 
w   o  6 

o  e 

Q. 

<U  _  £  ^ 
0 

<u 

■4-^  MIC 

a  V  1.3 


1  - 
0.5- 
0 


- 

/  ^  ^V-:- Cur-Sun 

2yr-Sun 

\ 

-  2yr-Shade 

■N   0 

- 

Cur-Shade 
—  '          '         '  ' 

M  J 

J       A       S  0 

2yr-Shade 


2yr-Sun 


M 


0 


1988 


Figure  5 — Seasonal  changes  in  net  photosyn- 
thetic  rates  {P^J  at  1 ,000  iimol/m^/s  and 
10  °C,  and  dark  respiration  rates  (fl)  at  10  °C 
of  P.  pumila  needles  during  1988.  □  =  current, 
sun-needles;  ■  =  current,  shade-needles; 
O  =  2-year-old,  sun-needles;  •  =  2-year-old, 
shade-needles. 


Needle  longevity  of  P.  pumila,  4  years  or  more,  corre- 
sponds to  that  of  P.  mugo  and  P.  contorta  (Benecke  and 
Havranek  1980). 

Climate  Factors  Affecting 
Photosynthetic  Production 

Seasonal  differences  in  photosynthetic  activity  between 
new  and  old  needles  have  been  reported  for  some  north- 
em  or  subalpine  evergreen  conifers  (Fry  and  Phillips 
1977;  Teskey  and  others  1984).  Likewise,  photos5Tithetic 
activity  (P^^J  in  the  2-year-old  needles  of  P.  pumila  be- 
came higher  between  late  July  and  early  August,  while 
that  of  current  needles  did  so  in  late  August  and  mid- 
September  (fig.  5).  A  similar  seasonal  trend  in  P^^  was 
found  for  the  current  and  older  needles  (1  to  4  years  old) 
of  P.  pumila  in  the  study  plot  in  1987  (Kajimoto  1990). 
These  results  suggest  that  photosynthetic  production  of 
the  P.  pumila  canopy  is  primarily  active  during  August 
and  September. 

Alpine  conifer  trees  rarely  achieve  their  potential  maxi- 
mum photosynthetic  activities  under  natural  environmen- 
tal conditions,  although  the  main  limiting  factors  in  photo- 
sythesis,  such  as  irradiance  and  air  and  soil  temperatures, 
seem  to  be  different  depending  on  species  and  site  condi- 
tions (De  Lucia  and  Smith  1987;  Hasler  1982;  Tiirner  and 
others  1983).  Mean  air  temperatures  between  late  July 
and  mid-September  at  the  research  plot  (fig.  1)  corre- 
sponded to  optimiun  air  temperatm-e  (10-15  °C)  for  net 
photosynthesis  of  P.  pumila,  both  for  current  and  old 


DISCUSSION 

Growth  Characteristics  of  Needles 

Needles  of  P.  pumila  began  to  elongate  in  early  July 
and  almost  finished  growth  in  late  August  (fig.  2).  The 
period  of  needle  growth  is  1  or  2  months  later  than  those 
of  other  low-altitude  pines,  for  example,  P.  densiflora 
(Tanaka  and  others  1976)  and  P.  thunbergii  (Nagatsu 
1987),  growing  naturally  in  central  Japan.  After  needle 
growth  had  terminated,  the  current  needles  of  P.  pumila 
showed  gradual  decline  in  specific  leaf  area  (SLA)  (fig.  3) 
and  maintained  higher  dark  respiration  rates  than 
2-year-old  needles  (fig.  5).  This  indicates  that  the  current 
needles  continued  to  accumulate  internal  carbohydrate  or 
nutrient  until  October. 

In  the  study  area,  there  was  little  local  variation  in  sea- 
sonal growth  patterns  in  the  shoots  and  needles  of  P. 
pumila  (Kajimoto  1989b).  However,  the  final  lengths  of 
current  shoots  and  needles  became  smaller  with  increas- 
ing altitude,  as  reported  for  other  high-altitude  pines, 
such  as  P.  cembra  (Baig  and  Tranquillini  1976),  and  P. 
mugo  and  P.  contorta  (Benecke  and  Havranek  1980). 

The  current  needles  of  P.  pumila  in  the  research  plot 
accounted  for  9  percent  of  total  foliage  biomass  in  late 
July  and  27  percent  in  late  August  (Kajimoto  1989a). 
This  indicates  that  70  percent  or  more  of  the  pine  canopy 
consisted  of  older  needles  throughout  the  growing  season. 


^  August 

^  30- 


u 
c 

Of 
D 
CT 
<1> 


> 

"a 
cr 


20 
10 


^  '0 


10 


September 

30 
20 
10 
0 


20 


Air  temperature  CO 


Figure  6 — Relative  frequency  distributions  of 
daytime  (8  a.m.  to  4  p.m.)  mean  air  tempera- 
ture at  each  1-hour  interval  in  August  (above) 
and  September  (below)  1 988.  Air  tempera- 
tures were  recorded  at  the  P.  pumila  stand 
(2,600  m  a.s.l.),  Kiso  Mountain  Range. 


96 


30- 


o 
c 

u 
cr 


20  - 
10  - 


August 


1000 


2000 


> 


40 


5  301- 
(T  20h 


September 


10  ^ 


0  *^^^^°"'f" 


0  1000  2000 

PFD      ( jjnnol  m'h''  ) 

Figure  7 — Relative  frequency  distributions  of 
daytime  (8  a.m.  to  4  p.m.)  photosyntiietic  photon 
flux  density  (PFD)  at  each  1  -hour  interval  in 
August  (above)  and  September  (below)  1 988. 
PFD  was  measured  above  the  canopy  surface  of 
P.  pumila  stand  (2,600  m  a.s.l.),  Kiso  Mountain 
Range. 

needles  (Kajimoto  1990).  Figure  6  shows  patterns  of  rela- 
tive frequency  distribution  of  daytime  mean  air  tempera- 
ture (8  a.m.  to  4  p.m.)  at  1-hour  intervals;  temperature 
values  between  10  and  16  °C  made  up  about  70  percent  of 
the  sum  total  in  both  August  and  September.  With  re- 
gard to  irradiance  conditions,  proportions  of  PFD  values 
above  600  |j.mol/m7s,  when  the  pine  needles  were  photo- 
synthetically  light  saturated  (table  1),  were  53  percent  in 
August  and  36  percent  in  September  (fig.  7).  The  tem- 
perature and  light  regimes  indicate  that  net  assimilation 
of  P.  pumila  needles  under  natural  conditions  is  limited 
mainly  by  lower  levels  of  solar  irradiance.  Reduction  in 
available  PFD  from  August  (53  percent)  to  September 
(36  percent)  is  likely  to  impede  net  photosynthesis  of  the 
current  needles,  since  photosynthetic  activities  (P^^)  of 
the  current  needles  were  the  highest  in  September  (fig.  5). 

It  is  concluded  from  the  present  study  that  P.  pumila 
needles  are  not  very  efficient  in  photosynthesis  diiring  the 
first  growing  season.  The  pine  needles  survive  for  four  or 
more  growing  seasons  and  recover  their  photosynthetic 
abilities  in  each  summer;  P     of  the  oldest  needles 

'  max 

(4  years  old)  was  about  half  of  that  of  1-year-old  needles, 
although  P^^  decreased  with  needle  age  (Kajimoto  1990). 
Consequently,  a  large  proportion  of  annual  net  carbon 
gain  in  the  P.  pumila  tree  is  likely  to  depend  on  net  as- 
similation by  the  old  needles. 

Knowledge  of  physiological  properties  of  P.  pumila, 
especially  water  relations  (Ando  and  Kawasaki  1991),  is 
still  limited  compared  to  other  subalpine  pine  species.  As 
to  seasonal  photosynthetic  production  of  P.  pumila,  more 
information  on  its  photosynthetic  reponses  to  environ- 
mental factors,  particularly  based  on  field  measurement, 
is  needed. 


ACKNOWLEDGMENTS 

I  thank  Dr.  K.  Hozumi  for  his  suggestions  and  en- 
couragement during  this  study,  and  Drs.  T.  Koike, 
M.  Ishizuka,  A.  Osawa,  Y.  Chiba,  and  N.  Kurachi  for  their 
useful  comments  on  this  manuscript.  Comments  by  Drs. 
F.-K.  Holtmeier,  W.  C.  Schmidt,  and  R.  Hasler  are  ac- 
knowledged. Thanks  also  go  to  Professors  T.  Horiuchi 
and  S.  Ito,  Shinshu  University,  and  to  the  staff  of  Shinshu 
University  Experimental  Forest  and  of  Nagoya  University 
Forest  for  their  assistance  during  the  field  siu"vey. 

REFERENCES 

Ando,  H.;  Kawasaki,  K.  1991.  Growth  of  Pinus  pumila  in 
relation  to  the  site  conditions  on  a  slope.  Transactions 
of  the  102d  Meeting,  Japanese  Forestry  Society: 
451-452.  [In  Japanese]. 

Baig,  M.  N.;  Tranquillini,  W.  1976.  Studies  on  upper  tim- 
ber line:  morphology  and  anatomy  of  Norway  spruce 
(Picea  abies)  and  stone  pine  (Pinus  cembra)  needles 
from  various  habitat  conditions.  Canadian  Journal  of 
Botany.  54:  1622-1632. 

Benecke,  V.;  Havranek,  W.  M.  1980.  Phenological  growth 
characteristics  of  trees  with  increasing  altitude, 
Craigieburn  Range,  New  Zealand.  In:  Benecke,  U.; 
Davis,  M.  R.,  eds.  Mountain  environments  and  subal- 
pine tree  growth.  Tech.  Pap.  70.  New  Zealand  Forest 
Service,  Forest  Research  Institute:  155-174. 

De  Lucia,  E.  H.;  Smith,  W.  K.  1987.  Air  and  soil  tempera- 
ture limitations  on  photosynthesis  in  Engelmann 
spruce  during  summer.  Canadian  Journal  of  Forest 
Research.  17:  527-533. 

Fry,  D.  J.;  Phillips,  I.  D.  J.  1977.  Photosynthesis  of  coni- 
fers in  relation  to  annual  growth  cycles  and  dry  matter 
production.  II.  Seasonal  photosynthetic  capacity  and 
mesophyll  ultrastructure  in  Abies  grandis,  Picea 
sitchensis,  Tsuga  heterophylla  and  Larix  leptolepis  in 
S.W.  England.  Physiologia  Plantarum.  40:  300-306. 

Hadley,  J.  L.;  Smith,  W.  K.  1987.  Influence  of  krummholz 
mat  microclimate  on  needle  physiology  and  survival. 
Oecologia.  73:  82-90. 

Hasler,  R.  1982.  Net  photosjTithesis  and  transpiration  of 
Pinus  montana  on  east  and  north  facing  slopes  at  al- 
pine timberline.  Oecologia.  54:  14-22. 

Holtmeier,  F.-K.  1973.  Geoecological  aspects  of  timber- 
lines  in  northern  and  central  Europe.  Arctic  and  Alpine 
Research.  5(3):  45-54. 

Holtmeier,  F.-K.  1981.  What  does  the  term  "krummholz" 
really  mean?  Observations  with  special  reference  to 
the  Alps  and  the  Colorado  Front  Range.  Mountain 
Research  and  Development.  1(3/4):  253-260. 

Kajimoto,  T.  1989a.  Aboveground  biomass  and  litterfall  of 
Pinus  pumila  scrubs  growing  on  the  Kiso  Mountain 
Range  in  central  Japan.  Ecological  Research.  4:  55-69. 

Kajimoto,  T.  1989b.  Current  growth  of  Pinus  pumila. 
Transactions  of  the  37th  Meeting,  Chubu  Branch, 
Japanese  Forestry  Society:  55-58.  [In  Japanese]. 

Kajimoto,  T.  1990.  Photosynthesis  and  respiration  of 
Pinus  pumila  needles  in  relation  to  needle  age  and 
season.  Ecological  Research.  5:  333-340. 


97 


Kajimoto,  T.  1992.  Dynamics  and  dry  matter  production 
of  belowground  woody  organs  of  Pinus  pumila  trees 
growing  on  the  Kiso  Mountain  Range  in  central  Japan. 
Ecological  Research.  7:  333-339. 

Larcher,  W.  1975.  Physiological  plant  ecology.  Heidelberg: 
Springer.  252  p. 

Nagano  Meteorological  Observatory.  1988.  Annual  report 
on  weather  conditions  of  Nagano  prefecture  in  1987. 
Nagano  Branch  Office  of  Japanese  Meteorological  Soci- 
ety. 137  p.  [In  Japanese]. 

Nagatsu,  M.  1987.  Correlation  between  the  growth  of 
needles  and  shoots  in  a  Pinus  thunbergii  yoimg  tree. 
Transactions  of  the  98th  Meeting  Japanese  Forestry 
Society:  417-418.  [In  Japanese]. 

Okitsu,  S.;  Ito,  K.  1984.  Vegetation  dynamics  of  the  Sibe- 
rian dwarf  pine  {Pinus  pumila  Regel)  in  the  Taisetsu 
Mountain  Range,  Hokkaido,  Japan.  Vegetatio.  58: 
105-113. 

Okitsu,  S.;  Ito,  K.  1989.  Conditions  for  the  development  of 

the  Pinus  pumila  zone  of  Hokkaido,  northern  Japan. 

Vegetatio.  84:  127-132. 
Sakai,  A.;  Larcher,  W.  1987.  Frost  survival  of  plants.  Ecol. 

Stud.  62.  New  York:  Springer.  321  p. 
Shidei,  T.  1963.  Productivity  of  Haimatsu  {Pinus  pumila) 

community  growing  in  alpine  zone  of  Tateyam-Range. 


Journal  of  the  Japanese  Forestry  Society.  45:  169-173. 
[In  Japanese  with  English  simimary]. 

Tanaka,  H.;  Oohata,  S.;  Akai,  T.  1976.  Elongation  and 
shoot  forms  in  foreign  pines.  Report,  Kyoto  University 
of  Forestry.  11:  38-49.  [In  Japanese]. 

Teskey,  R.  O.;  Grier,  C.  C;  Hinckley,  T.  M.  1984.  Changes 
in  photosynthesis  and  water  relations  with  age  and  sea- 
son in  Abies  amabilis.  Canadian  Journal  of  Forest 
Research.  14:  77-84. 

Tranquillini,  W.  1979.  Physiological  ecology  of  the  alpine 
timberline.  Ecol.  Stud.  31.  New  York:  Springer.  131  p. 

Turner,  H.;  Hasler,  R.;  Schonenberger,  W.  1983.  Contrast- 
ing microenvironments  and  their  effects  on  carbon  up- 
take and  allocation  by  young  conifers  near  alpine 
treeline  in  Switzerland.  In:  Wareing,  R.  H.,  ed.  Carbon 
uptake  and  allocation  in  subalpine  ecosystems  as  a  key 
to  management:  Proceedings  of  lUFRO  workshop.  1982 
August  2-3.  Corvallis,  OR:  Oregon  State  University, 
Forest  Research  Laboratory:  22-30. 

Wardle,  P.  1977.  Japanese  timberlines  and  some  geo- 
graphic comparisons.  Arctic  and  Alpine  Research.  9: 
249-258. 


98 


HEIGHT  GROWTH  IN  CEMBRAN  PINE 
AS  A  FACTOR  OF  AIR  TEMPERATURE 

Herbert  Kronfuss 


Abstract — The  study  treats  height  growth  in  cembran  pine 
iPinus  cembra)  in  high-elevation  afforestation  at  an  altitude 
of  1,800  m  with  growth  patterns  analyzed  as  a  factor  of  air  tem- 
perature. It  is  assumed  that  growth  behavior  follows  an  endog- 
enous rhythm  possibly  deriving  from  the  genotj^je.  The  relation- 
ship between  growth  pattern  and  air  temperature  is  demon- 
strated for  defined  periods  and  phases  of  height  growth.  From 
the  results  it  becomes  obvious  that  height  growth  and  thus  the 
relative  rate  of  growth  are  primarily  dependent  on  temperature. 
Data  from  long-term  monitoring  series  were  used  to  determine 
the  ampUtudes  between  the  growth  curves  for  early  and  for  late 
culmination.  A  comparison  of  the  ampUtudes  for  cembran  pine 
with  spruce  and  larch  shows  cembran  pine — with  the  highest 
amplitudes — to  be  the  most  sensitive  to  temperature.  Cembran 
pine  thus  makes  most  efficient  use  of  warm  weather  periods  for 
increment  production  in  short  periods  of  time. 


The  cembran  pine  (Pinus  cembra)  occupies  the  highest 
stands  of  all  trees  in  the  Alps  and  as  such  is  ideally  suited 
for  a  study  of  height  growth  related  to  air  temperature. 
The  fact  that  cembran  pine  is  not  thought  to  have  differ- 
ent ecotypes  adapted  to  various  altitudes  also  permits 
growth  characteristics  of  this  species  to  be  defined  in  spe- 
cific terms. 

STUDY  SITE 

The  study  site  was  located  30  km  southwest  of  Innsbruck, 
Tyrol  (Austria),  in  the  Stubai  Alps  near  Haggen  in  the 
Sellrain  Valley.  For  location,  geology,  and  climate  details, 
see  figures  1  and  2,  and  table  1.  The  results  relate  to  a 
continuous  experimental  stand  of  cembran  pine  at  an  alti- 
tude of  1,800  m.  The  stand  was  established  in  1970  when 
the  plot  was  afforested  with  4-year-old  plants. 

MONITORING 

Monitoring  was  performed  over  a  period  of  8  years 
for  a  group  of  40  trees.  Height  growth  in  cembran  pine 
was  measured  at  5-day  intervals  (pentades)  and  the  rate 
of  growth  correlated  with  mean  air  temperature  of  the 
pentade.  Individual  height  increments  were  expressed 
as  a  percentage  of  annual  cumvdative  grov^^h  to  permit 
the  growth  patterns  to  be  compared  -with,  changes  in 
temperatxire. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Herbert  Kronfuss  is  Forester,  Forestry  Research  Institute  of  Austria, 
Rennweg  1/29,  A-6020  Innsbruck. 


Growth  patterns  are  represented  in  the  form  of  relative 
growth-rate  c;irves  and  relative  cumulative-growth  curves. 
This  not  only  permits  direct  comparisons  to  be  made  be- 
tween annual  increments  but  also,  given  a  long  enough 
monitoring  period,  permits  a  characteristic  increment  pat- 
tern (type)  to  be  derived  for  each  species  (Burger  1926; 
Larcher  1980;  Week  1955). 

HEIGHT  GROWTH 

From  1978  to  1985  increment  growth  lasted  an  average 
of  108  days,  vnth  a  20-day  variance  range. 

To  study  the  influence  of  temperature  on  the  growth 
pattern  during  the  period  of  extension  growth,  the  rela- 
tive grovrth-rate  curve  was  correlated  with  the  tempera- 
ture curve,  with  the  mean  height  increments  from  con- 
secutive pentades  expressed  as  a  percentage  of  total 
height  growth.  The  relative  growth-rate  curve  clearly 
illustrates  the  pattern  of  height  grov^h  (fig.  3). 

Figure  3  shows  how  temperature  fluctuations  during 
the  period  of  extension  growth  influence  the  rate  of  height 
growth.  With  regard  to  the  correlation  between  the 
height-growth  behavior  and  temperattire  patterns,  the 
following  points  would  seem  to  be  of  interest. 

Dming  growing  season,  the  temperature  curve  itself 
is  characterized  by  two  pronotmced  dips  for  the  periods 
May  20  to  25  and  Jtme  15  to  20,  a  phenomenon  well 
knovm  to  the  local  rural  communities  as  the  "Ice  Saints" 
and  "Sheep's  Chill."  At  the  beginning  of  April,  mean  tem- 
perattire  reaches  freezing  point  on  the  long-term  average. 
The  rise  in  temperature  continues  imtil  mid-May,  reach- 
ing an  average  of  approximately  7  °C,  before  dropping 
back  to  5  °C  for  the  pentade  occupying  the  first  of  the  two 


Figure  1 — Geographical  location  of  the  site. 


99 


HAGGEN  i.S.dSOO  m)  a)  3,1°  b)  920  mm 
c)  10         d)  1975  -  1984 


mm 
200 

100 


10 


I  I  I  I  I 


I  I  I  I  I  I  I 


I  I  I  I  I 


M 


M 


a)  Mean  annual  temperature 

b)  Mean  annual  precipitation 

c)  Number  of  years  of  data  collection  for  temperature  and 
precipitation 

d)  Period  of  data  collection  for  temperature  and 
precipitation:  1975-1984 

e)  Mean  monthly  precipitation 

f)  Mean  monthly  temperature 

g)  Months  with  average  daily  minimum  below  0°C 

h)  Months  with  absolute  minimum  temperatures  below  0°C 

Figure  2 — Data  for  Haggen  Climate  Diagram 
Haggen  I.S.,  Field  Station  1, 1,800  m,  SSW. 


cold  spells.  During  the  time  of  the  "Ice  Saints,"  most  of 
the  cembran  pines  are  at  the  beginning  of  the  period  of  ex- 
tension growth,  with  a  growth  rate  of  only  2  to  3  percent 
of  relative  growth  per  pentade.  That  corresponds  to  about 
1.5  cm.  On  the  average,  approximately  7  percent  of  cumu- 
lative growth  is  achieved  by  this  point  in  time.  For  height 
grovi^h  in  cembran  pines  in  this  location,  the  period  be- 
tween the  "Ice  Saints"  and  "Sheep's  Chill"  is  of  special 
importance  as  it  includes  the  first  major  peak  in  the  tem- 
perature curve,  on  or  around  May  31  on  average. 

With  a  brief  delay,  this  temperature  peak  stimulates 
a  response  in  the  cembran  pine  in  terms  of  height  growth, 
and  after  one-  and  one-half  pentades,  the  biggest  absolute 
increment  is  produced,  accounting  for  about  35  percent  of 


total  growth  on  a  long-term  average.  Following  this  peak 
in  increment  production,  the  typical  pattern  of  height 
growth  in  the  cembran  pine  is  one  of  declining  rates  of 
growth  in  spite  of  rising  temperatures.  This  is  shown  in 
figure  3  with  the  two  curves  moving  in  opposite  directions 
following  the  start  of  the  "Sheep's  Chill"  period  aroimd 
June  20. 

Not  even  an  increase  in  mean  pentade  temperature 
beyond  the  10  °C  mark  around  July  5,  finally  peaking 
at  13  °C,  leads  to  a  further  spontaneous  spate  of  growth. 


Table  1 — Location  of  the  study:  HAGGEN  afforestation  site  in  the 
Sellrain  Valley  near  St.  Sigmund,  Tyrol 


Geographical  coordinates: 
Height  above  sea-level: 
Exposition  and  angle  of  slope: 


47°13'N.  ir06'E. 
1,715-1,950  m 
SSW,  30° 


Bedrock:         Foliated  gneiss  (biotite  granite  gneiss,  granodiorite 
gneiss)  and  mica  schist. 

Vegetation  at 

afforestation:  Heath  (callunetum,  nardetum  alpigenum) 
Soils:  Humus  horizons  developed  into  a  brown  earth  type 

through  grazing  and  haymaking.  More  or  less 

podsolic  brown  soils. 
In  previous  centuries  a  heavily  grazed  avalanche  slope. 

CLIMATIC  DATA  are  based  on  a  10-year  monitoring  period  (1975- 
1984) 

-  Sunshine  hours  during  the  vegetation  period  (mid-May  to 

September) 

a)  theoretical  asatronomical  maximum  2,005  hours 

b)  theoretical  local  maximum  1 ,393  hours 
horizon  loss                                         30.5  percent 

-  Air  temperature 

a)  mean  annual  temperature  3.1  °C 

b)  during  vegetation  period  8.4  °C 

-  Precipitation 

a)  mean  annual  precipitation  (3  m  aboveground)         920  mm 

b)  during  vegetation  period  (3  m  aboveground)  533  mm 

c)  precipitation  at  ground  level  650  mm 
(=  approx.  22  percent  higher  than  at 

3  m  aboveground) 

-  Precipitation  probability  (percent) 

a)  June  maximum  55  percent 

b)  December  minimum  27  percent 

-  Mean  precipitation  distribution  (mm) 

a)  maximum  in  July  and  September  8.4  mm 

b)  minimum  in  February  3.6  mm 

c)  month  with  highest  precipitation  (July)  141  mm 

d)  month  with  lowest  precipitation  (February)  33  mm 

-  Evaporation  during  the  vegetation  period 

a)  potential  evaporation  (2.5  m  aboveground  566  mm 
with  "piche")  (=  4.1  l/m^  per  day) 

b)  actual  evapotranspiration  (lysimeter  with  236  mm 
ground  vegetation  (=  1 .7  l/m^  per  day) 

c)  actual  hydrological  budget  412  mm 
(=  3  l/m^  per  day) 

The  meteorological  summer  semester  accounts  for  65  percent  of 
annual  precipitation  (598  mm),  and  the  meteorological  winter 
semester  35  percent  (322  mm). 


100 


Mean  pentad  values  for  relative  growth  rate 
and  air  temperature  (°C) 


Mean  relative  growth  rate  (%) 


January  February    March     April        May      June       July     August  September  October  November  December 

Figure  3 — Comparison  between  the  curves  of  relative  growth  rate  and  air 
temperature,  8-year  series  (1978  to  1985). 


The  minor  fluctuations  in  the  growth  curve  on  the  order 
of  1  to  3  percent  can  be  correlated  with  temperature  fluc- 
tuations of  10  to  13  °C. 

The  main  features  of  the  curves  in  figure  3  can  be  de- 
scribed as  follows:  The  grovvi;h  curve  for  cembran  pine 
peaks  (maximum  growth  rate  per  pentade  is  reached) 
on  June  5  to  10  on  average.  The  temperature  curves  are 
characterized  by  two  peaks,  the  first  between  the  two  pro- 
nounced dips  (May  31  to  June  5)  and  the  second  between 
July  25  and  July  31.  It  is  interesting  to  note  that  height 
increments  between  July  5  and  the  end  of  the  period  of 
extension  growth — and  that  is  a  good  46  days — total 
only  20  percent  of  cumulative  grov\i;h,  even  though  the 
temperature  curve  passes  10  °C  on  July  5  and  remains 
more  or  less  at  that  level  until  mid-September. 

The  above  picture,  based  on  the  mean  figures  calculated 
from  monitored  data,  illustrates  in  bold  strokes  the  corre- 
lation between  the  temperature  and  height-growth  patterns. 

The  relative  grov^i:h-rate  curves  shown  in  figure  4  for 
the  8  years  from  1978  to  1985  provide  an  overview  of 
growth  patterns  in  cembran  pine  as  a  factor  of  air  tem- 
perature for  the  individual  years.  One  of  the  more  strik- 
ing features  of  these  curves  is  the  variation  in  the  onset 
of  culmination  (equals  greatest  increment  production  per 
pentade).  On  this  basis  we  can  distinguish  between  an- 
nual growth  curves  with  early  culmination  (1979,  1981, 
1982,  1983),  one  year  with  late  culmination  (1984),  and 
years  with  twin  or  intermediate  peaks  (1978,  1980,  1985; 
see  also  table  2). 

A  comparison  of  the  temperature  curve  in  figure  3  with. 
the  growth-rate  curves  in  figure  4  reveals  two  clearly 
delineated  time  periods  in  which  culmination  (peak 
growth  rate  per  pentade)  occurs.  The  first  is  the  period 
between  the  "Ice  Saints"  and  "Sheep's  Chill,"  and  the  sec- 
ond is  a  subsequent  period  of  30  days  ending  approxi- 
mately with  the  pentade  of  July  15  to  20.  These  two  peri- 
ods together  total  57  days  (equals  53  percent  of  the  mean 


 1978 

...z 

■  1  

. . . .  1  

!^^^7^  ^ 

10  20  31   10  20  30  10  20  31  10  20  31 

May  June      July  August 


W  20  31  10  20  30  10  20  31  10  20  31 


—  »« 

J. 

■   

^^~.yrr>^  

10  20  31   10  20  30  10  20  31  10  20  31 


 1980 

■  ■1  



W  20  31   10  20  30  10  20  31  10  20  31 


 1981 

■A^  

«  20  31   10  20  30  10  20  31  10  20  31 


 1983 

,,l  



10  20  31   10  20  30  10  20  31   10  20  31 


10  20  31   10  20  30  10  20  31   10  20  31 


 1986 

10  20  31  10  20  30  10  20  31  10  20  31 


Figure  4 — Cembran  pine  relative  height-growth 
curves  (percent). 


101 


Table  2 — Air  temperature  in  °C  and  percentage  growth  in  cembran 
pine' 


J   n  I  I 


Year 

TS 

T 

TS 

T 

HG 

HG 

 °C- 

-  -  Percent  -  - 

Early  culmination 

1979 

245 

9.1 

298 

9.9 

71 

22 

1981 

242 

8.9 

230 

7.7 

61 

19 

1982 

246 

9.1 

368 

12.3 

52 

O  IT 

35 

1983 

203 

7.5 

359 

12.0 

46 

38 

Average 

234 

8.7 

314 

10.5 

58 

28 

Intermediate  culmination 

8.2 

241 

8.0 

38 

52 

1980 

195 

7.2 

205 

6.8 

26 

47 

1985 

191 

7.1 

284 

9.5 

31 

50 

Average 

203 

7.5 

243 

8.1 

32 

50 

Late  culmination 

1984 

169 

6.3 

284 

9.5 

17 

58 

'I  =  period  from  the  "Ice  Saints"  to  "Sheep's  Chill";  II  =  period  from  "Sheep's 
Chill"  to  20  July;  TS  =  temperature  sum;  T  =  mean  temperature;  HG  =  height 
growth;  period  I  lasts  for  27  days;  period  II  lasts  for  30  days. 


10      20         31         10         20         30         10         20        31         10        20  31 

May  June  July  August 


Figure  5 — Cembran  pine  relative  height  growth 
(percent). 


period  of  extension  growth)  and  account  for  an  average 
of  82  percent  of  annual  shoot  length. 

There  is  a  significant  correlation  between  percentage 
growth  within  the  two  main  growth  periods  and  tempera- 
ture sum,  v^dth  peak  growth  located  in  the  period  that  ex- 
hibits the  highest  temperature  sum  in  absolute  terms, 


except  that  beyond  an  average  temperature  sum  of  230  °C 
for  the  first  period  a  further  increase  in  temperature  for 
the  second  period  does  not  trigger  higher  growth  rates  in 
that  period.  The  highest  pentade  growth  rate  always  oc- 
curs in  the  period  with  the  highest  temperature  Bum  and 
coincides  with  or  follows  the  pentade  with  maximum 
pentade  temperature  (see  table  2). 

From  table  2  it  can  be  seen  that  in  the  case  of  early  cul- 
mination the  average  temperature  sum  in  the  first  period 
was  234  °C  (8.7  °C/27  days).  During  that  period,  an  aver- 
age of  58  percent  of  extension  growth  was  measured  (with 
scatter  between  46  percent  and  71  percent).  In  those 
cases  the  cembran  pines  achieved  an  average  of  only  28 
percent  relative  growth  in  the  second  period  in  spite  of  the 
rise  in  the  average  temperature  sum  to  314  °C  (10.5  °C/30 
days). 

The  year  of  late  culmination,  1984,  is  also  instructive. 
The  cold  weather  in  period  I  of  that  year  produced  a  tem- 
perature sum  of  only  169  °C  (6.3  °C/27  days),  which  is 
reflected  in  a  correspondingly  low  relative  growth  rate 
of  17  percent.  Period  II,  however,  was  characterized 
by  a  significant  rise  in  temperature,  producing  a  tem- 
perature sum  of  284  °C  (9.5  °C/30  days).  The  relative 
growth  rate  in  that  period  (fig.  5)  was  58  percent,  which 
is  identical  with  the  figure  for  period  I  in  the  years  of 
early  culmination. 

At  the  end  of  July  in  the  late  culmination  year, 
the  leaders  had  reached  90  percent  of  total  extension 
growth  compEired  with  98  percent  for  the  years  of  early 
culmination. 

This  analysis  of  the  height-growth  patterns  of  cembran 
pine  over  an  8-year  period  of  observation  (1978  to  1985) 
shows  that  the  cembran  pine  is  ontogenetically  equipped 
to  make  efficient  use  of  available  warmth  for  increment 
production  so  as  to  achieve  ctdmination  as  early  as  pos- 
sible. Up  to  that  point,  air  temperature  is  the  dominant 
factor  in  determining  the  relative  growth  rate,  £ind  after- 
ward it  continues  to  function  as  the  "engine"  that  main- 
tains the  extension  process. 

PHASES  OF  HEIGHT  GROWTH 

The  following  analysis  of  annual  height  growth  by  help 
of  the  growth  curve  (Kronfuss  1985)  will  illustrate  the 
somewhat  abstract  correlation  between  growth  rate  and 
temperature  with  the  aid  of  further  figures.  For  this  pur- 
pose, height  growth  can  be  divided  into  the  following 
three  phases  (fig.  6): 

a.  Initial  phase  (0  to  25  percent  of  annual  extension). 

b.  Main  phase  (25  to  75  percent  of  annued  extension). 

c.  Final  phase  (75  to  100  percent  of  annual  extension). 

Table  3  provides  a  separate  assessment  of  the  correla- 
tion for  each  of  these  three  phases  using  a  regression  line, 
correlation  coefficient,  and  regression  coefficient  to  ex- 
press the  relationship  between  mean  daily  temperature 
(T/d)  and  average  height  growth  per  day  (HG/d)  (fig.  6, 
table  3). 

Whereas  the  correlation  coefficient  for  the  initial  phase 
(r  =  0.6)  suggests  a  correlation  vnth  temperature,  this  is 
no  longer  the  case  in  the  final  phase  (r  =  0.42).  The  main 


102 


Correlation  between  mean  daily  air  temperature  and  height  increment 
per  day  HAGGEN  i.S.,  1800  m 

Average  height  increment  per  day  (mm) 


4- 

r  =  0.89  k=^^!£A 

^      Mam  phase 

Initial  phase 

r=.0.42  /f.O.II 

• 

•  • 

r=o.60  ;f=o.is  ■ 

•  •• 

 1  i_ 

*           "  *  Final  phase 

_      1  ■ 

S           6             7  8 

9  10 

It             12           13  14 

Air  temperature  ("C) 


Correlation  between  mean  daily  air  temperature  and  growth  processes 
HAGGEN  i.S..  1800  m 

Duration  of  phases  (days) 


Air  temperature  (°C) 

Figure  6 — Cembran  pine  correlations  between 
mean  daily  air  temperatures,  height  increment,  and 
duration  of  growth  phases:  period  1 978  to  1 985. 


growth  phase,  on  the  other  hand,  is  largely  determined  by 
temperature,  as  can  be  seen  from  the  almost  stretched 
curve  (fig.  6  at  top)  and  a  correlation  coefficient  of  r  =  0.89. 

The  regression  coefficients  {h),  as  the  expression  of  aver- 
age daily  growth  per  vinit  of  temperature,  show  that,  for 
the  same  temperature  rise,  height  growth  in  the  main 


phase  is  4.5  times  greater  than  in  the  initial  phase  and 
6.2  times  greater  than  in  the  final  phase.  In  the  initial 
phase,  with  an  average  period  of  extension  growth  of  32 
days,  the  average  daily  increment  is  1.6  mm.  In  the 
main  growth  phase  the  period  of  extension  growth  is  re- 
duced to  an  average  of  26  days  of  an  HG/d  of  4  mm,  while 
the  figure  for  the  subsequent  50  days  of  the  final  phase  is 
only  1  mm. 

The  duration  of  extension  growth  also  correlates  with 
temperature  and  is  shorter  with  higher  temperature  rise. 
The  influence  of  temperature  on  the  period  of  extension 
growth  can  be  seen  from  the  correlation  coefficients,  with 
a  close  correlation  with  temperature  expressed  by  the  co- 
efficients r  =  0.82  for  the  initial  phase  and  r  =  0.84  for 
the  main  growth  period,  decreasing  in  the  final  phase 
to  r  =  0.59.  This  low  correlation  coefficient  shows  that 
reduced  height  growth  in  the  final  phase  is  no  longer  due 
exclusively  to  the  influence  of  temperature,  but  to  the 
tailing  away  of  an  endogenous  pattern.  With  regard  to 
the  temperature-dominated  phases  of  extension  growth, 
(the  initial  and  main  phases)  an  increase  in  mean  tem- 
perature by  1  °C  reduces  the  extension  growth  period  by 
4  to  5  days.  Also,  in  the  final  phase  we  see  again  that  the 
pattern  derives  from  an  autonomous  endogenous  rhythm 
that  determines  the  potential  period  of  extension  growth 
in  this  phase  for  the  species  involved. 

GROWTH  CURVE  VARIANTS 

As  can  be  seen  from  the  shapes  of  the  relative  growth- 
rate  curves  in  figure  4,  annual  patterns  of  height  growth 
vary  considerably  for  the  same  species  in  the  same  loca- 
tions. Eight  years  of  observation  offer  enough  data  for 
graphic  representation  of  maximum  amplitudes  between 
the  growth  curves  for  the  earliest  and  those  for  the  latest 
culminations  and  for  a  comparison  with  spruce  and  larch 
(fig.  7).  This  illustrates  not  only  the  time  sequences  with 
regard  to  flushing  but  also  the  cxirrent  range  of  fluctua- 
tion for  these  tree  species. 


Table  3 — Correlations  between  mean  daily  temperature  (t7d)  and  height  growth  per  day  (HG/d)  and  between  mean  dally 
temperature  (t7d)  and  the  period  of  extension  growth  days  (d) 


Year 

Initial  phase^ 

Main  phase^ 

Final  phase^ 

t°/d 

HG/d 

d 

t°/d 

HG/d 

d 

t°/d 

HG/d 

d 

1978 

7.9 

1.8 

24 

6.9 

2.6 

33 

10.4 

0.9 

48 

1979 

9.6 

1.8 

22 

8.8 

4.7 

17 

9.9 

.6 

71 

1989 

5.4 

.9 

43 

6.6 

2.5 

32 

12.0 

1.0 

42 

1981 

6.5 

1.7 

25 

.4 

4.7 

18 

8.9 

.7 

59 

1982 

8.2 

1.6 

28 

9.3 

4.1 

22 

11.7 

.9 

52 

1983 

6.1 

1.8 

33 

8.8 

4.6 

26 

13.1 

1.5 

40 

1984 

5.2 

1.3 

52 

9.7 

5.5 

25 

9.8 

1.6 

42 

1985 

7.5 

2.1 

29 

7.2 

3.5 

35 

11.2 

1.4 

43 

Average 

7.1 

1.6 

32 

8.5 

4.0 

26 

10.9 

1.1 

50 

'Initial  phase:  correlation  coefficient  r=  0.604;  regression  coefficient  /( =  0.147;  point  of  intersection  b=  0.587;  standard  deviation  s=  0.326. 
^Main  phase:  r  =  0.889;  k  =  0.681 ;  b  =  -1 .471 ;  s  =  1 .000. 
^Final  phase:  r  =  0.426;  /c  =  0. 1 1 7;  b  =  -0. 1 95;  s  =  0.353. 


103 


Cumulative  growth  (%)  / 

 ^  -   -.--^ 

—  ^  V/- 

/   0  / 
  /    cr    y  / 

/-/"  /•  

/ '''' 

tj~r" 

r  1  1  1 

1     1     L     1     1     1     J     1     1     1     J     1     1  1 

1     10    20   30    10    20    31    10    20    30    10    20    31    10    20    31    10    20  30 
April  Sept. 

Figure  7 — Characteristic  growth  curve  ampli- 
tudes for  cembran  pine,  spruce,  and  European 
larch:  afforestation  "Haggen"  in  1 ,800  m  a.s.l. 


Apart  from  the  differences  in  the  beginning  and  end  of 
the  period  of  extension  growth,  cembran  pine  is  the  most 
sensitive  to  temperature  of  the  three  species,  as  the  high 
amplitude  shows.  It  can  be  concluded  that  the  cembran 
pine  makes  the  most  efficient  use  of  warm  weather  condi- 
tions for  increment  production.  In  the  initial  phase  of 


extension  growth,  cembran  pine  is  already  well  ahead  of 
the  other  species  in  terms  of  growth.  In  the  main  growth 
phase,  larch  is  also  surpassed  by  the  cembran  pine,  while 
spruce  comes  very  close  to  it. 

For  the  main  growth  period,  the  two  extremes  of  the 
growth  curves  for  spruce  and  larch  are  almost  parallel. 
This  shows  that  increment  production  per  pentade  is 
more  balanced  between  warm  and  cold  conditions  than 
in  the  case  of  cembran  pine,  which  is  much  more  sensitive 
to  temperature  fluctuations. 

REFERENCES 

Burger,  H.  1926.  Untersuchungen  iiber  das  Hohen- 
wachstum  verschiedener  Holzarten.  Mitt.  Schweiz. 
Centralanst.  Forstl.  Vers,  wesen;  XIV/I. 

Kronfuss,  H.  1985.  Die  Zuwachsleistimg  einer  Hochlagen- 
aufforstung  mit  Zirbe  (Pinus  cembra)  auf  einem 
Siidhang,  in  Abhaangigkeit  von  der  Seehohe.  lUFRO 
Meeting  Proceedings.  Davos,  Switzerland. 

Larcher,  W.  1980.  Okologie  der  Pflanzen  auf 
physiologischer  Grundlage.  Ulmer,  Stuttgart. 

Week,  J.  1955.  Forstliche  Zuwachs  und  Ertragskimde. 
Neimiann,  Berlin. 


104 


SURVIVAL  AND  GROWTH  OF  PLANTED 
CEMBRAN  PINES  AT  THE  ALPINE 
TIMBERLINE 


Josef  Senn 

Walter  Schonenberger 
Ueli  Wasem 


Abstract — In  an  experimental  afforestation  with  cembran  pines 
{Pinus  cembra  L.)  at  the  subalpine  timberhne,  tree  survival  was 
mainly  determined  by  the  date  of  disappearance  of  the  snow 
cover  in  spring.  Trees  survived  well  on  early  snow-free  sites  and 
poorly  on  sites  with  long-lasting  snow  cover.  The  major  causes  of 
mortality  were  the  two  parasitic  fungi  Ascocalyx  abietina  and 
Phacidium  infestans.  Tree  height,  as  an  index  for  growth,  was 
dependent  on  the  amount  of  energy  available  during  the  vegeta- 
tion season,  and  therefore  negatively  correlated  with  both  alti- 
tude above  sea  level  and  wind  velocity,  and  positively  with  global 
radiation. 


In  the  severe  winter  of  1951-52  avalanches  caused 
major  destruction  all  over  the  Alps.  This  initiated  a 
multidisciplinary  research  program  to  restore  the  upper 
treeline.  The  aim  of  the  program  was  to  develop  methods 
to  alforest  treeless  avalanche  catchments  and  abandoned 
meadows  within  the  subalpine  forest  belt  in  order  to  es- 
tablish new  protection  forests. 

Protection  forests  are  an  important  feature  in  alpine 
landscapes.  Many  areas  can  only  be  permanently  inhab- 
ited as  long  as  these  forests  remain  intact.  In  some  of 
these  forests,  however,  severe  problems  concerning  stand 
stability  have  become  evident.  Inadequate  plantation 
techniques  in  the  past  have  frequently  led  to  forests  that 
were  highly  susceptible  to  natural  hazards  such  as 
storms.  On  the  other  hand,  locally  dense  ungulate  popu- 
lations prevented  successful  natiu-al  rejuvenation  in  the 
existing  forests. 

Both  natural  regeneration  and  artificial  afforestation 
generally  have  become  more  difficult  toward  the  alpine 
timberline,  since  all  biological  processes  become  slower 
with  increasing  altitude  above  sea  level.  In  general,  com- 
pared to  natiiral  invasion  of  forest  trees,  planting  reduces 
the  time  needed  to  establish  forests  that  fulfill  their  pro- 
tective function.  Therefore  artificial  afforestation  may  be 
highly  desirable.  Poor  financial  resources,  however,  may 
Umit  planting  to  the  most  promising  sites. 

The  experimental  afforestation  at  StiUberg  was  designed 
to  obtain  information  about  site  conditions  that  determine 
success  or  failure  of  future  afforestations,  and  to  be  able 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Josef  Senn,  Walter  Schonenberger,  and  Ueli  Wasem  are  members  of  the 
High  Altitude  Afforestation  Group  at  the  Swiss  Federal  Institute  for  For- 
est, Snow  and  Landscape  Research,  CH-8903  Birmensdorf,  Switzerland. 


to  inform  forestry  personnel  which  sites  would  be  most 
promising  for  successful  planting  and  on  which  sites 
planting  would  be  fruitless. 

Cembran  pine  {Pinus  cembra  L.)  is  one  of  the  three  tim- 
berline species  tested  at  StiUberg.  This  species,  extending 
to  the  highest  altitudes  of  any  trees  in  the  Alps,  played  an 
important  role  in  the  culture  of  local  human  populations 
(Holtmeier,  these  proceedings),  produces  valuable  timber, 
and  is  esteemed  by  "green"  tourism  as  an  important  fea- 
ture of  natural  landscapes.  Nonsustainable  exploitation 
in  medieval  times  and  pasturalism  reduced  cembran  pine 
to  a  fraction  of  its  original  abundance,  although  there  is 
evidence  for  natural  range  expansion  in  the  past  100  years. 
Artificial  reproduction  and  estabhshment  of  cembran  pine, 
on  the  other  hand,  pose  many  problems  for  practical  for- 
estry (Frehner  and  Schonenberger,  these  proceedings). 

MATERIAL  AND  METHODS 

The  present  study  was  conducted  at  StiUberg  experi- 
mental area,  a  highly  structiu-ed  northeast  slope  (fig.  1) 
in  the  Dischma  Valley  near  Davos,  Switzerland  (lat. 
34°08'  S.,  long.  171°41'  E.).  The  area  extends  over  9.9  ha 
between  2,080  and  2,230  m  above  sea  level.  The  highest 
parts  reach  beyond  the  altitudinal  timberline.  Before 
plantation  in  1975  various  site  aspects  like  slope,  expo- 
sure, global  radiation  (Turner  1966),  wind  conditions 
(Nageli  1971),  date  of  disappearance  of  snow  cover  in 
spring  (Rychetnik  1987),  ntunber  of  snow-fi-ee  days  in 
winter,  avalanche  fi'equencies,  soil  types  (Blaser  1980), 
and  vegetation  (mainly  dwarf  shrub  communities)  (Kuoch 
1970)  were  recorded  on  a  very  fine  scale.  Schonenberger 
(1975)  described  the  site  in  detail.  Results  from  the 
StiUberg  experimental  afforestation  were  published  by 
Schonenberger  and  others  (1988).  Practical  aspects  for 
high-altitude  forestry  were  included  in  Schonenberger 
and  others  (1990). 

Three  coniferous  tree  species  were  planted  in  1975: 
5-year-old  cembran  pines,  3-year-old  mountain  pines  (P. 
uncinata  (Miller)  Domin  =  Pinus  montana  Miller  =  Pinus 
mugo  Turra  ssp.  uncinata  Domin)  the  erect  form  of  the 
mugo  pine,  and  1-year-old  European  Isirches  (Larix  de- 
cidua  Miller).  Before  planting,  the  whole  area  was  di- 
vided into  square  units  of  3.5  by  3.5  m  (fig.  1).  In  each 
square  unit  25  trees  of  one  species  were  planted.  The 
number  of  trees  per  imit  may  have  been  reduced  by  ad- 
verse local  soil  conditions  like  barren  rock.  The  squares 
contained  alternatively  cembran  pines,  mountain  pines, 
and  larches.  Altogether  some  90,000  trees  were  planted 
in  4,052  square  units. 


105 


Figure  1 — Topographic  model  of  the  Stillberg 
experimental  area  with  the  grid  of  the  affores- 
tation units  (3.5  by  3.5  m). 


This  paper  reports  the  results  of  survival  of  the  cembran 
pines  in  comparison  with  mountain  pine  and  European 
larch  in  the  whole  study  area,  and  of  height  variation  and 
mortality  causes  in  a  subsample.  Survival  of  each  tree 
was  recorded  annually  as  from  1976  (number  of  square 
units  with  cembran  pines  =  1,351,  with  mountain  pines  = 
1,350,  and  with  larches  =  1,351).  Mean  tree  height,  gen- 
eral constitution,  type  and  degree  of  abiotic  and  biotic 
damage,  and  causes  of  mortality  were  determined  annu- 
ally in  228  square  units  containing  cembran  pines,  in  226 
square  imits  containing  mountain  pines,  and  in  226 
square  units  containing  larches.  Twenty-eight  qualitative 
"types"  of  damage  and  potential  causes  of  death  could  be 
attributed  in  the  field. 

Correlation  analysis  was  used  to  test  for  interactions 
between  various  site  parameters  and  tree  performance. 
The  relative  importance  of  site  parameters  was  evaluated 
by  stepwise  regression  analyses  (SAS  1985)  for  survival 
and  tree  height.  The  survival  rates  were  arcsine  trans- 
formed before  performing  stepwise  regression  analysis. 

RESULTS 
Survival 

During  the  first  3  years  after  planting,  survival  was 
highest  in  cembran  pine  compared  to  mountain  pine  and 
larch,  but  in  1991  only  25.3  percent  of  the  planted  cembran 
pines  survived,  compared  to  37.8  percent  in  mountain 
pine  and  73.1  percent  in  larch  (fig.  2). 

Stepwise  regression  of  survival  rate  in  a  square  unit 
(number  of  trees  alive  in  1991/number  of  trees  planted  in 
1975)  of  cembran  pine  revealed  that  the  date  of  disappear- 
ance of  snow  cover  in  spring  and  the  altitude  above  sea 
level  were  the  most  important  factors,  explaining  some 
46  percent  and  8  percent  of  total  variation,  respectively. 
In  our  study  area  the  survival  rate  was  high  on  the  early 
snow-free  and  lower  sites.  Survival  decreased  with  in- 
creasing altitude,  and  almost  no  trees  survived  on  sites 
where  snow  lasted  on  10-year  average  longer  than  June  10. 


Avalanche  frequency,  slope  inclination,  global  radiation, 
and  the  number  of  snow-free  days  in  winter  also  had  sig- 
nificant effects  on  survival,  but  each  of  these  factors  con- 
tributed less  than  1  percent  to  the  total  variation. 

Mortality  Causes 

Of  the  28  recorded  causes  for  mortality,  Ascocalyx 
able-  tina  (Lagerb.)  (=  Gremmeniella  abietina  [Lagerb.] 
Morelet),  a  parasitic  fungus,  was  the  major  factor,  killing 
some  52.3  percent  of  the  planted  cembran  pines.  A  second 
fungus,  Phacidium  infestans  Karst.,  the  snow  blight, 
killed  another  11.8  percent  of  the  trees.  Some  4.8  percent 
of  the  trees  died  or  disappeared  without  any  detectable 
reason.  Undetermined  fungi  killed  1.3  percent  and  com- 
petition with  herbaceous  vegetation  some  0.6  percent  of 
the  planted  trees.  Herbivores  caused  no  losses. 

We  found  a  significantly  positive  correlation  between  the 
proportion  of  cembran  pines  killed  by  Ascocalyx  abietina 
and  the  date  of  disappearance  of  snow  cover  in  spring 
(fig.  3).  Square  units  that  were  snow  free  before  May  12 
suffered  only  slightly  (13.9  percent  of  the  trees  killed), 
whereas  AscocaZyx  killed  90.3  percent  of  the  trees  in  units 
that  were  covered  by  snow  later  than  June  10.  Occur- 
rence oi Ascocalyx  was  also  positively  correlated  with 
altitude  {r  =  0.55,  P  <  0.0001,  n  =  228),  the  fungus  kilHng 
an  increasing  proportion  of  yoimg  trees  with  increasing 
altitude. 

Between  the  proportion  of  cembran  pines  killed  by  Pha- 
cidium infestans  and  the  date  of  disappearance  of  snow 
cover  in  spring,  however,  we  found  a  negative  correlation 
(r  =  -0.42,  P  <  0.001,  n  =  228).  This  means  that  the  high- 
est proportion  of  trees  was  killed  in  squares  that  were 


75  76  77  78  79  80  81  82  83  84  85  86  87  88  89  90  91 


Year 

Figure  2— Survivorship  curves  of  the  three 
tree  species,  cembran  pine  (Pinus  cembra), 
mountain  pine  (P.  uncinata),  and  European 
larch  {Larix  decidua),  planted  in  1975  in  the 
Stillberg  experimental  area. 


106 


1.0 


p 

E  o 

O  0) 

a 
o 


0.8  • 


0.6 


0.4 


0.2 


•  .  .  •  •  • 


•  •  •       I  • 


•i 

•t  . 


•     •  • 

t'..'  «. 


120  130  140  150  160  170 

Date  of  Disappearance  of  Snow  Cover  (day) 

Figure  3 — Relationship  between  the  propor- 
tion of  cembran  pines  killed  in  a  square  unit  by 
Ascocalyx  abietina  (=  number  of  trees  killed  by 
Ascocalyx  since  1 976/number  of  trees  planted 
in  1975)  and  the  date  of  disappearance  of  snow 
cover  in  spring.  Day  1 30  =  May  1 0;  day  1 60  = 
June  9.  Regression  line:  y  =  -3.0637  +  0.0251  x, 
r=0.8^,P«  0.0001,  n  =  228. 


snow  free  early,  whereas  the  importance  of  Phacidium 
as  a  mortaHty  cause  decreased  with  increasing  duration 
of  snow  cover.  Similarly  the  ratio  of  trees  killed  by  Pha- 
cidium decreased  with  increasing  altitude  {r  =  —0.26, 
P<0.0001,  n  =  228). 

The  occurrences  of  the  two  parasitic  fungi  were  signifi- 
cantly negatively  correlated  within  square  units  (r  =  -0.59, 
P  <  0.0001,  n  =  228).  If  one  of  the  two  fungus  species  was 
common  in  one  square,  the  other  was  rare. 


DISCUSSION 
Survival 

Duration  of  snow  cover  in  spring  was  found  to  be  the 
main  limiting  factor  for  survival  of  young  cembran  pines 
in  the  upper  subalpine  forest  belt.  On  sites  where  snow 
cover  lasted  until  after  Jime  10  cembran  pines  were  al- 
most completely  excluded.  The  second  evergreen  Pinus 
species,  the  mountain  pine,  was  similarly  vulnerable  to 
long-lasting  snow  cover.  The  European  larch,  a  deciduous 
conifer,  was  much  more  tolerant.  In  this  species  30  per- 
cent of  the  trees  siu^ved  on  sites  snow  covered  until  after 
June  10.  Further,  survival  of  cembran  pine  was  signifi- 
cantly affected  by  altitude,  although  this  factor  was  clearly 
less  important  than  duration  of  snow  cover  in  spring.  In- 
terestingly, the  impact  of  site  factors  such  as  global  radia- 
tion, number  of  snow-free  days,  slope  inclination,  and  ava- 
lanche frequency  on  tree  survival  was  negligible  compared 
to  the  first  two  factors.  Avalanche  frequency,  however, 
will  certainly  become  more  important  in  the  near  future. 
Young  trees  with  flexible  stems  generally  survive  ava- 
lanche impacts  without  significant  damage.  But  with  in- 
creasing height  and  stem  diameter  the  trees  are  becoming 
increasingly  vulnerable  to  stem  breakage  caused  by  snow 
movements,  a  tendency  that  has  become  evident  only  dur- 
ing the  last  years. 

Mortality  Causes 

In  our  study  area  the  two  parasitic  fungi  Ascocalyx 
abietina  and  Phacidium  infestans  were  the  major  mortal- 
ity factors  for  cembran  pines,  and  these  fungi  still  kill  sig- 
nificant numbers  of  young  trees  every  year. 


Tree  Height 

Mean  tree  height  in  1991  was  lowest  in  cembran  pine 
(58.8  cm)  compared  to  moxmtain  pine  (72.7  cm)  and  larch 
(59.8  cm),  although  initial  mean  height  was  largest  in 
cembran  pine  (fig.  4). 

Variation  in  height  was  considerable,  depending  on 
local  site  conditions.  On  sunny  southeast  exposed  sites, 
average  height  of  cembran  pines  was  above  80  cm,  whereas 
on  shady  sites  mean  height  was  around  45  cm.  Stepwise 
regression  revealed  that  tree  height  was  mainly  affected 
by  altitude,  global  radiation,  and  wind  velocity,  these  fac- 
tors contributing  21  percent,  15  percent,  and  10  percent 
to  the  total  variation.  Duration  of  snow  cover  in  spring 
explained  3  percent  of  the  variation  in  height.  In  our 
study  area  the  cembran  pines  were  high  on  sites  at  low 
altitudes,  receiving  high  amounts  of  radiation  and  being 
protected  from  strong  winds.  Slope  inclination,  avalanche 
frequency,  and  number  of  snow-free  days  in  winter  had  no 
significant  impact  on  tree  height. 


E 

o 


c 

(0 
0) 


80 


70 


60 


SO 


X 


30 


20 


10 


1 

-♦■  - 

-  C€ 

-  M< 

-  La 

>mb 
}unt 
rch 

'an  pine 
ain  pine 

r' 

r' 

— 

-«-•• 

r' 

,  

✓ 

r' 

r'' 

b-' 



✓ 

...I 

p: — 

r' 

— yt 
(' 

1  1 



— 

r-- 

r-- 

r 

..-1 

r' 

r'' 

r 



1  .- 

■ 

75  76  77  78  79  80  81  82  83  84  85  86  87  88  89  90  91 

Year 

Figure  4 — Annual  mean  height  of  the  three  tree 
species,  cembran  pine  {Pinus  cembra),  mountain 
pine  (P.  uncinata),  and  European  larch  (Larix  de- 
cidua),  planted  in  1975  in  the  Stillberg  experimen- 
tal area. 


107 


Ascocalyx  infestations  first  become  evident  in  spring 
in  the  dying  buds.  Later  in  the  season  the  shoots  die  back 
along  the  branches.  The  fungus  may  eventually  kill  a 
young  cembran  pine  after  2  or  3  years.  The  close  relation- 
ship between  the  date  of  disappearance  of  snow  cover  in 
spring  and  the  occurrence    Ascocalyx  (fig.  3)  indicates 
that  the  fungus  preferentially  attacks  weakened  plants. 
On  the  sites  with  long-lasting  snow  cover,  the  vegetation 
season  is  relatively  short,  and  thick  layers  of  raw  himius 
have  accumulated  because  of  the  low  temperatures. 
Therefore,  they  are  poor  sites  for  tree  growth.  Similarly, 
Kurkela  (1984)  found  that  the  trees  most  heavily  affected 
hy  Ascocalyx  had  the  lowest  growth  rate  before  the  epi- 
demic. He  explained  the  differences  in  tree  growth  by  the 
different  site  fertility.  In  our  study  the  hypothesis  that 
Ascocalyx  preferentially  attacked  weakened  trees  was  fur- 
ther supported  by  the  finding  that  the  proportion  of  trees 
killed  by  the  fungus  significantly  increased  with  increas- 
ing altitude  (toward  the  altitudinal  tree  line).  In  our  study 
area  Ascocalyx  attacked  mountain  pines,  too,  where  it  also 
caused  considerable  losses. 

In  contrast  to  the  situation  in  Ascocalyx,  the  proportion 
of  cembran  pines  killed  by  Phacidium  was  negatively  cor- 
related with  the  date  of  disappearance  of  snow  cover  and 
with  the  altitude.  According  to  this  correlation  the  trees 
growing  at  lower  altitudes  and  on  sites  that  were  snow 
free  early  were  killed  more  frequently  by  Phacidium  than 
the  trees  growing  at  higher  altitudes  and  on  sites  with 
long-lasting  snow  cover.  Young  trees  that  looked  healthy 
in  one  year  were  often  found  dead  in  the  following  spring. 
Interestingly,  Roll-Hansen  and  others  (1992)  reported 
from  Scandinavia  that  the  most  vigorously  growing  plants 
were  most  susceptible  to  Phacidium  infestations,  whereas 
poorly  developing  plants  were  most  resistant. 

The  fact  that  Phacidium  killed  the  highest  proportion 
of  trees  at  early  snow-free  sites  seems  contradictory  to 
the  biology  of  the  fungus,  that  exclusively  attacks  parts 
of  trees  that  are  covered  by  snow  (Roll-Hansen  1989). 
Branches  destroyed  by  this  fungus  are  clearly  visible  after 
disappearance  of  snow  cover  and  mark  the  prevalent  snow 
depth.  Our  study  area  contains  some  older  cembran  pines 
growing  on  ridges  (at  the  sites  that  are  snow  free  early). 
These  trees  originating  from  caches  of  the  nutcracker 
{Nucifraga  caryocatactes)  host  Phacidium,  which  is  non- 
lethal  for  the  large  trees  that  extend  considerably  beyond 
the  snow  cover.  These  trees  may  act  as  sources  for  local 
Phacidium  outbreaks  that  eventually  kill  the  young  trees 
completely  covered  by  snow.  Since  this  fungus  specifically 
attacks  cembran  pine  and  does  not  occur  on  mountain 
pine  (but  see  Roll-Hansen  1989),  it  may  not  spread  as  eas- 
ily over  larger  areas  as  does  Ascocalyx,  which  uses  moun- 
tain pine  as  an  alternative  host.  Therefore,  even  heavier 
outbreaks  of  Phacidium  remain  locally  restricted  around 
mature  cembran  pines.  Together  with  the  fact  that  Pha- 
cidium preferentially  attacks  the  most  vigorous  trees,  this 
may  explain  the  seemingly  paradoxical  negative  correla- 
tion between  duration  of  snow  cover  in  spring  and  occur- 
rence o{  Phacidium. 

On  sites  with  high  avalanche  frequencies  the  young 
trees  were  partially  excluded,  not  by  stem  breakage  or 
uprooting,  but  by  competitive  interactions  with  herba- 
ceous vegetation.  The  moving  snow  carried  rich  mineral 


soil  leading  to  a  lush  herbaceous  vegetation  that  out- 
competed  the  slowly  growing  young  cembran  pines.  Ava- 
lanches, however,  become  a  more  serious  problem  for  the 
surviving  trees  as  they  increase  in  height  and  stem  diam- 
eter. Broken  stems  were  found  more  frequently  in  recent 
years. 

Herbivore  impact  on  the  survival  of  young  cembran  pines 
was  relatively  unimportant.  Herbivores  killed  no  trees,  al- 
though they  may  reduce  tree  growth.  Browsing  vertebrates 
like  black  grouse  (Tetrao  tetrix)  and  chamois  (Rupicapra 
rupicapra),  however,  clearly  preferred  mountain  pine  and 
larch  to  cembran  pine.  Black  grouse  caused  some  locally 
restricted  damage  by  browsing  buds  and  first-year  needles. 
They  obviously  preferred  to  browse  on  easily  accessible 
trees  growing  on  ridges  with  little  snow  cover  (Streule 
1973).  Populations  of  herbivorous  insects  like  phloem- 
sucking  aphids  (Cinara  cembrae  Seitner,  see  Grbic,  these 
proceedings,  and  Pineus  cembrae  [Choi.]  Amand.)  fluctu- 
ated in  size  but  were  low  in  most  of  the  years  when  com- 
pared to  the  other  two  tree  species.  Although  visible  dam- 
age by  herbivores  was  relatively  rare,  much  rarer  than 
fungal  infections,  the  herbivores  may  have  indirectly  af- 
fected the  survival  of  the  trees  by  injuring  plant  tissue, 
making  the  trees  more  susceptible  to  infestations  by  fungi. 

Tree  Height 

Tree  height  in  cembran  pine  was  clearly  dependent  on 
the  available  amount  of  energy  during  the  vegetation  sea- 
son. Altitude  above  sea  level,  global  radiation,  and  wind 
velocity  explained  the  largest  amount  of  variation  in 
height.  We  are  aware  that  tree  height  does  not  directly 
represent  tree  growth,  since  we  did  not  measure  annual 
growth.  We  only  recorded  tree  height  in  summer,  exclud- 
ing the  growing  shoot  of  the  current  year.  Tree  height, 
which  is  the  result  of  annual  growth  minus  the  losses 
through  damage  in  the  apical  region,  may  increase  or 
decrease  from  one  year  to  another.  Sixteen  years  after 
plantation,  differences  in  average  tree  height  among  sites, 
however,  are  primarily  the  result  of  variation  in  growth 
caused  by  variation  in  local  site  conditions. 

The  date  of  disappearance  of  snow  cover  in  spring,  al- 
though significantly  related  to  tree  height,  explained  only 
3  percent  of  the  variation  in  height  in  contrast  to  survival 
where  the  snow  conditions  in  spring  mainly  explained  the 
observed  pattern. 

Slope  inclination  and  avalanche  frequency,  which  may 
be  related  to  stem  breakage,  had  no  detectable  impact  on 
tree  height.  Further,  we  found  no  relationship  between 
the  number  of  snow-free  days  in  winter  and  tree  height, 
although  trees  that  were  snow  free  during  extended  peri- 
ods in  winter  should  have  been  more  vulnerable  to  herbi- 
vore browsing  than  trees  that  were  covered  by  deep  snow. 

CONCLUSIONS 

Foresters  should  carefully  select  favorable  microsites 
when  planting  cembran  pines  in  high-altitude  afforestations. 
Particular  attention  should  be  paid  to  the  spatial  pattern 
of  snowmelt  in  spring,  and  no  cembran  pines  should  be 
planted  at  sites  with  prolonged  duration  of  snow  cover. 


108 


REFERENCES 

Blaser,  P.  1980.  Der  Boden  als  Standortsfaktor  bei 
Aufforstiingen  in  der  subalpinen  Stufe  (Stillberg, 
Davos).  Eidgenossische  Anstalt  fur  das  forstliche 
Versuchswesen,  Mitteilimgen.  56(3):  527-611. 

Kuoch,  R.  1970.  Die  Vegetation  auf  Stillberg  (Dischmatal, 
Kt.  Graubiinden).  Eidgenossische  Anstalt  fiir  das 
forstliche  Versuchswesen,  Mitteilungen.  46(4):  329-342. 

Kurkela,  T.  1984.  The  growth  of  trees  affected  by  Grem- 
meniella  abietina.  In:  Manion,  P.  D.,  ed.  Scleroderris 
canker  of  conifers.  The  Hague:  Martinus  NijhofiEDr  W. 
Junk  Pubhshers:  177-180. 

Nageh,  W.  1971.  Der  Wind  als  Standortsfaktor  bei 
Aufforstungen  in  der  subalpinen  Stufe  (Stillbergalp 
im  Dischmatal,  Kanton  Graubiinden).  Eidgenossische 
Anstalt  fiir  das  forstliche  Versuchswesen,  Mitteilungen. 
47(2):  33-147. 

Roll-Hansen,  F.  1989.  Phacidium  infestans:  a  literature 
review.  European  Journal  of  Forest  Pathology.  19: 
237-250. 

Roll-Hansen,  F.;  Roll-Hansen,  H.;  Skroppa,  T.  1992. 
Gremmeniella  abietina,  Phacidium  infestans,  and  other 
causes  of  damage  in  alpine,  young  pine  plantations  in 
Norway.  European  Journal  of  Forest  Pathology.  22: 
77-94. 

Rychetnik,  J.  1987.  Snow  cover  disappearance  as  influ- 
enced by  site  conditions,  snow  distribution,  and  ava- 
lanche activity.  In:  Proceedings  of  the  international 
symposiima  on  avalanche  formation,  movement  and 
effects;  Davos  1986.  lAHS-Pubhcation  Nr.  162. 

SAS.  1985.  SAS  user's  guide:  statistics.  Gary,  NC:  SAS 
Institute. 


Schonenberger,  W.  1975.  Standorteinfliisse  auf 
Versuchsaufforstimgen  an  der  alpinen  Waldgrenze 
(Stillberg,  Davos).  Eidgenossische  Anstalt  fiir  das 
forstliche  Versuchswesen,  Mitteilungen.  51(4):  357-428. 

Schonenberger,  W.  1985.  Performance  of  a  high  altitude 
afforestation  under  various  site  conditions.  In:  T\irner, 
H.;  Tranquillini,  W.,  eds.  Establishment  and  tending  of 
subalpine  forest:  research  and  management:  Proceed- 
ings of  the  3rd  lUFRO  workshop  P  1.07-00,  1984. 
Eidgenossische  Anstalt  fiir  das  forstUche  Versuchswesen, 
Berichte.  270:  233-240. 

Schonenberger,  W.;  Frey,  W.,  eds.  1988.  Untersuchimgen 
zur  Okologie  und  Technik  der  Hochlagenaufforstung; 
Forschungsergebnisse  aus  dem  Lawinenanrissgebiet 
Stillberg.  Schweizerische  Zeitschrift  fiir  Forstwesen. 
139:  735-820. 

Schonenberger,  W.;  Frey,  W.;  Leuenberger,  F.  1990. 
Okologie  und  Technik  der  Aufforstung  im  Gebirge  - 
Anregimgen  fiir  die  Praxis.  Eidgenossische  Anstalt  fiir 
das  forstliche  Versuchswesen,  Berichte  Nr.  325.  58  p. 
[Version  franfaise:  Ecologie  et  technique  des  afforesta- 
tions en  montagne  -  Suggestions  a  I'usage  des  praticiens. 
Traduction  du  rapport  No.  325.]  [Versione  italiana: 
Ecologia  del  rimboschimento  di  montagna  e  pratica 
della  tecnica  di  piantagione.  Traduzione  del  rapporto 
No.  325.] 

Streule,  A.  1973.  Schaden  in  Gebirgsaufforstungen  durch 
das  Birkhuhn  (Lyrurus  tetrix).  Biindnerwald.  28(8): 
249-254. 

Tiimer,  H.  1966.  Die  globale  Hangbestrahlung  als 
Standortsfaktor  in  der  subalpine  St\ife.  Eidgenossische 
Anstalt  fiir  das  forstliche  Versuchswesen,  Mitteilungen. 
42(3):  111-168. 


109 


Influence  of 
Environmental  Factors 


International  Workshop 
St.  Moritz  1 992 

s   ^' 


110 


CHANGES  OF  SWISS  STONE  PINE 
APHID  LIFE  CYCLE,  DENSITY,  AND 
POPULATION  STRUCTURE  IN  HIGH- 
ALTITUDE  SWISS  STONE  PINE 
AFFORESTATION 

Mihailo  Grbic 


Abstract — The  order  of  form  appearance,  structure,  and  number 
of  individuals  in  colonies  of  Cinara  cembrae  on  Swiss  stone  pine 
{Pinus  cembra)  used  for  afforestation  in  the  Dischma  Valley  were 
strongly  related  to  the  slope  aspect.  The  abnormality  was  par- 
ticularly apparent  on  north  slopes.   North  slope  colonies 
showed  incomplete  life  cycles.  All  insects  seen  there  were 
virginoparae  and  their  larvae.  Colonies  were  small  in  number 
and  were  developed  from  winged  virginoparae  from  other  areas. 


The  Alps  are  of  great  importance  to  Switzerland  as  an 
area  for  hviman  habitation  and  as  a  health  and  holiday  re- 
sort. Afforestation  in  high  altitudes  is  costly  and  involves 
high  risk,  as  many  failures  occur.  Because  of  the  slow  de- 
velopment of  such  afforestation  it  is  difficult  to  under- 
stand and  to  determine  the  causes  and  processes  that  lead 
to  a  failure  and  then  to  suggest  measures  to  improve 
success. 

Environmental  factors  of  the  upper  timberline  area  are 
totally  different  from  those  of  the  forest  zones.  These 
harsh  conditions  affect  the  specific  growth  and  develop- 
ment of  planted  species;  for  example,  low  survival  rate 
and  low  annual  increment.  The  surviving  plants,  depend- 
ing on  the  biological  characteristics  of  the  species,  are 
more  or  less  open  to  insect  and  fungal  attack.  These  same 
high-altitude  factors  (for  example,  low  temperature,  short 
growing  period,  and  short  day  photoperiod)  tend  to  change 
the  life  cycle,  density,  and  population  structure  of  insects. 
This  is  particularly  true  for  aphids  due  to  their  polymor- 
phic nature,  and  their  ability  to  alter  the  nimaber  of  gen- 
erations per  season. 

STUDY  AREAS  AND  METHODS 

These  investigations  were  carried  out  in  field  plots  of 
the  Stillberg  research  area  and  the  Lucksalp  comparative 
afforestation  area,  in  the  timberline  area  of  the  Dischma 
Valley  (Canton  Graubunden).  The  period  of  the  study 
was  between  late  May  and  the  middle  of  September, 
which  is  the  time  of  greatest  insect  activity. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Mihailo  Grbic  is  Teacher  at  the  University  of  Belgrade,  Faculty  of  For- 
estry, Kneza  Viseslava  1,  YU-11030  Belgrade,  Yugoslavia. 


Stillberg  Hes  at  2,080  to  2,230  m  above  sea  level  (a.s.l.), 
with  a  northeast  aspect,  and  30°  to  45°  slope.  The  slope  is 
divided  by  spurs,  and  as  a  result  there  are  three  different 
aspects  (north-,  northeast-,  and  east-facing  slopes).  On 
the  opposite  side  of  Dischma  Valley  lies  Lucksalp  at 
2,200  m  a.s.l.,  with  a  southwest  aspect. 

The  subjects  of  investigation  were  19-year-old  Swiss 
stone  pine  (Pinus  cembra  L.)  and  the  Swiss  stone  pine 
aphid  (Cinara  cembrae  Seithner),  the  most  frequently  ob- 
served insect  pest  on  that  afforestation  species.  The  trees 
were  planted  70  cm  apart  in  a  grid  pattern  in  approxi- 
mately 4,000  square  plots  (3.5  by  3.5  m).  The  Swiss  stone 
pine  stocks  were  alternated  from  plot  to  plot  with  two 
other  afforestation  species,  17-year-old  Swiss  mountain 
pine  (Pinus  mugo  ssp.  uncinata  [Mill.]  Domin.),  and 
15-year-old  European  larch  (Larix  decidua  Mill.),  each 
plot  containing  25  trees  (Shonenberger  1985). 

The  samples  of  Swiss  stone  pine  aphid  were  identified 
by  comparing  them  to  recent  accounts  from  various  parts 
of  Europe  (Carter  and  Maslen  1982;  Eastop  1972;  Pintera 
1966;  Stroyan  1955,  1960).  The  structure  of  antennae, 
rostrum,  abdomen,  and  hind  legs  of  collected  material 
were  compared  with  these  references. 

The  life  cycle  was  observed  in  colonies  feeding  on  par- 
ticular sample  trees.  On  Lucksalp,  30  trees  (5  trees  x 
6  plots)  were  regtdarly  examined,  and  50  (5  trees  x 
10  plots)  were  examined  on  Stillberg.  Observations  were 
made  to  establish  the  relationship  between  the  intensity 
of  infestation  and  site  aspects. 

The  trees  were  examined  every  10  days  to  determine 
density  of  pest  population  and  structvire  of  colonies. 

MORPHOLOGY 

The  morphological  characteristics  of  specimens  found 
on  Stillberg  and  Lucksalp  (tables  1  and  2)  are  similar  to 
the  central  European  description  (Pintera  1966).  The 
chronological  series  of  morphs  is:  fundatrix,  aptera 
virginopara,  alata  virginopara,  ovipara,  and  aptera  and 
alata  males. 

Compared  with  the  apterae  of  the  succeeding  genera- 
tions, fundatrix  shows  "fundatrix  facies"  characteristics 
(Lees  1961).  With  the  Swiss  stone  pine  aphid,  the  anten- 
nae and  legs  are  relatively  shorter  in  comparison  to  the 
body.  The  processus  terminalis  is  shorter  with  fewer 
rhinaria.  Features  similar  to  the  fundatrix  facies  that 
were  observed  by  Hille  Ris  Lambers  (1955),  Lees  (1961), 
and  Stroyan  (1960)  on  nonfimdatrix  morphs  of  some 


111 


Table  1 — Biometric  data  for  different  morphs  of  Cinara  cembrae  from  Lucksalp 


Biometric  data 


Aptarae 
virginoparae 


Alatea 
virginoparae 


Morph 
Apterae 
male 


Alatae 
male 


Sexual 
female 


Body  length  (mm) 

Length  of  antennal  segments 
II 
IV 
V 

VI  (base  +  proces.  termlnalls) 

Length  of  rostal  segments  (^) 
IV 
V 

Hind  tarsus  segments  (^i) 
I 

basal  diameter 
dorsal  length 
ventral  length 

II 

length 

Hind  tibia  length  (mm) 

Length  of  longest  hair  on 
III  ant.  segment 
hind  tibia 

VIII  abdom.  tergite 

Number  of  hairs  on 
II  ant.  segment 
VI  ant.  segment 
base 

proces.  term, 
ultimate  rostral  segment 

Number  of  hairs  on 
subgenital  plate 
VIII  abdom.  tergite 

Number  of  secondary  rhinaria  on 
ant.  segments 

III 

IV 

VI 


3.05-3.52 


680-720 
300-340 
400-410 
167-180  + 
75-83 

218-238 
100-112 


74-  78 
95-105 

155-182 

385-420 
3.20-3.30 

75-  97 
75-120 

110-165 

9-15 

9-  13 
4-5 
6-8 

10-  12 
4-6 


0-2 
0-2 
2-3 


3.00-3.10 


635-650 
275-290 
320-345 
150-160  + 
59-66 

170-182 
85-88 


57-62 
62-66 
122-130 

330-350 
2.90-3.05 

88-92 
70-105 
185-200 

6-7 

10-12 
3-4 
8 

10 


9-11 
2-3 
5 


2.50-3.30 


830-880 
335-400 
430-500 
190-198  + 
75-90 

230-250 
90-110 


60-70 
98-110 
150-180 

340-390 
1.85-3.00 

75-120 
75-110 
100-112 

10-15 

10-12 
5 

10-12 

10-12 
7-11 


10-13 
7-10 
4-5 


2.40-3.00 


845-910 
360-430 
440-500 
109-198  + 
85-95 

220-230 
105-110 


60-65 
85-105 
150-180 

385-400 
2.90-3.40 

115-140 
140-180 
90-120 

7-  14 

8-  10 

5-  6 
8-10 

6-  9 
6-9 


55-83 
10-13 
5 


3.80-4.70 


830-890 
360-385 
480-500 
192-212  + 
75-80 

245-250 
106-118 


68-88 
105-120 
187-200 

370-420 
3.70-3.78 

95-123 
112-145 
120-138 

10-16 

8-  13 
5-6 

9-  11 

9-13 
7-8 


2-3 
1-2 
4-7 


aphids  that  live  at  low  temperatures  were  not  recorded  in 
these  study  areas. 

Morphological  differences  between  apterous  and  alatae 
virginoparae  are  found  not  only  in  the  presence  or  ab- 
sence of  the  wings,  but  also  in  the  following  characteris- 
tics. In  alatae: 

•  Segmentation  between  head  and  prothorEix  is  more 
conspicuous. 

•  Compound  eyes  are  bigger,  v(ath  many  facets. 

•  Antennae  bear  five  times  as  many  rhinaria. 

•  The  body  is  shorter,  and  the  mesothoracic  terga  con- 
sists of  a  compound  notum  and  postnotum. 

•  The  scutum  occupies  the  central  major  part  of  the 
nottun  and  is  divided  in  two  mesothoracic  lobes,  which  are 
hardly  developed. 


The  virginoparae  of  every  generation  are  morphologi- 
cally uniform,  and  under  field  conditions  it  is  impossible 
to  find  out  how  many  generations  have  developed.  Also, 
the  length  of  time  required  for  passing  through  four  in- 
stars,  from  birth  to  adult,  is  variable  and  dependent  on 
two  external  factors  (food  quality  and  temperature)  and 
two  internal  factors  (birth  weight  and  whether  the  morph 
is  winged  or  unwdnged)  (Dixon  1987).  This  is  the  reason 
for  the  presence  of  different  generations  of  the  same  form 
occurring  simultaneously  in  one  colony. 

As  is  known,  Swiss  stone  pine  aphids  produce  both 
alatae  and  apterous  males.  Both  forms  were  found  in  the 
research  areas,  but  with  a  preponderance  of  the  wingless 
form.  Both  forms  of  males  are  readily  recognized  by  their 
sclerotized  genital  structure.  The  body,  especially  abdo- 
men, is  smaller  and  more  slender,  and  the  antennae  bear 
more  rhinaria. 


112 


Table  2 — Biometric  data  for  different  morphs  of  Cinara  cembrae  from  Stillberg 


Morph 


Aptarae 

Alatea 

Apterae 

Alatae 

Sexual 

Biometric  data 

virginoparae 

virginoparae 

male 

male 

female 

Body  length  (mm) 

2.90-3.70 

2.90-3.90 

2.50-3.10 

2.40-3.00 

3.90-4.70 

Length  of  antennal  segments 

II 
II 

540-800 

^  Art    rt  C  rt 

630-650 

820-800 

840-930 

800-860 

IV 

260-355 

250-280 

320-400 

360-440 

350-400 

V 

340-450 

www  OOU 

tOU-wuU 

ARn-t^  1  n 

HOU  0  1  u 

u-ouu 

VI  (base  +  proces.  termlnalis) 

170-188-1- 

153-167  + 

183-210  + 

192-200  + 

175-222  + 

80-85 

58-68 

75-95 

88-95 

75-85 

Length  of  rostal  segments  (n) 

IV 

rtOQ  OCA 

238-250 

1 62-1 72 

235-250 

220-225 

235-250 

V 

100-108 

83-87 

95-102 

100-105 

105-118 

Hind  tarsus  segments  (n) 
1 

h^^^sl  dif^mptAr 

75-78 

WW  w9 

DU-DO 

DO- 1 UO 

dorsal  length 

88-108 

60-62 

95-1 02 

88-1 07 

105-125 

ventral  length 

II 

158-175 

122-127 

150-178 

150-170 

185-230 

length 

365-385 

300-340 

335-395 

380-400 

420-450 

Hind  tibia  length  (mm) 

2.40-3.39 

2.82-2.98 

1 .78-2.98 

2.95-3.37 

3.70-3.88 

Length  of  longest  hair  on 

III  ant.  segment 

60-00 

85-88 

78-125 

1 1 2-1 42 

95-130 

hind  tibia 

60-105 

75-1 10 

78-1 12 

140-175 

1 12-150 

VIII  aooom.  tergite 

168-220 

195-210 

100-115 

88-130 

112-138 

Number  of  hairs  on 

II  ant.  segment 

11-15 

6-7 

10-16 

7-15 

10-18 

VI  ant.  segment 

base 

8-12 

9-13 

10-13 

8-1 1 

8-16 

proces.  term. 

4-5 

3-4 

5-6 

5 

4-6 

ultimate  rostral  segment 

8-9 

7-9 

9-12 

8-11 

8-11 

Number  of  hairs  on 

subgenital  plate 

9-11 

6 

9-12 

6-9 

9-14 

VIII  abdom.  tergite 

4-6 

7-12 

6-9 

6-9 

Number  of  secondary  rhinaria  on 

ant.  segments 

III 

0-4 

8-10 

10-15 

50-89 

2-4 

IV 

2-3 

1-3 

7-11 

11-15 

0-2 

VI 

3-4 

6 

4-6 

5-6 

2-7 

Alatae  and  apterous  males  and  alatae  virginoparae  pos- 
sess a  far  larger  number  of  secondary  rhinaria  than  apter- 
ous virginoparae  and  sexuales  females.  The  following 
numbers  of  secondary  rhinaria  on  the  third  antennal  seg- 
ment were  found:  apterous  virginoparae  0-4,  alatae 
virginoparae  8-11,  apterous  males  10-15,  and  alatae 
males  50-89.  The  number  of  secondary  rhinaria  on 
sexuales  females  was  similar  to  that  found  in  apterous 
virginoparae. 

The  function  of  secondary  rhinaria  is  probably  different 
in  male  and  female  individuals.  The  secondary  rhinaria 
in  males  appear  to  be  the  main  receptors  of  sex  phero- 
mones,  while  secondary  rhinaria  of  alatae  virginoparae 
must  have  a  function  other  than  pheromone  detection. 
This  function  is  possibly  host  selection,  as  has  been  shown 
by  many  authors  with  other  aphid  species  (Alikhan  1960; 
Chapman  and  others  1981;  and  Pospisil  1976). 


Sexuales  females  resemble  the  apterous  virginoparae, 
but  may  be  somewhat  bigger.  The  tibiae  of  the  hind  legs 
are  often  longer  with  longer  setae  also,  and  numerovis 
pseudosensoria  on  them. 

LIFE  CYCLE 

General — The  life  cycle  of  the  Swiss  stone  pine  aphid  is 
strongly  related  to  site  aspects  and  environmental  factors. 
Generally,  the  Swiss  stone  pine  aphid  has  a  monoecious 
holocyclic  type  of  life  cycle.  The  life  cycle  starts  in  May 
v^dth  the  hatching  of  fundatrices  larvae  from  eggs  laid  in 
rows  on  the  needles.  In  the  covirse  of  svmimer  a  small 
niunber  of  generations  of  parthenogenetic  females 
develop. 

Generations  of  winged  parthenogenetic  viviparous  fe- 
males arise  from  colonies  of  apterae  later  in  the  summer. 


113 


Sexuales  forms  appear  in  the  second  half  of  August.  This 
early  development  of  sexuales  is  due  to  the  alpine  climate. 
In  the  low-lying  positions  where  Swiss  stone  pine  has 
been  artificially  introduced,  Swiss  stone  pine  aphid  has 
never  been  found. 

Observed — The  life  cycle  observations  in  the  Dischma 
Valley  show  the  following  characteristics:  colonies  from 
Lucksalp  and  all  slopes  of  Stillberg,  except  north-facing 
slopes,  show  complete  life  cycles  with  six  morphs.  Par- 
allel appearance  of  two  forms  of  males,  observed  by 
Seithner  (1936),  is  a  unique  case  in  the  family  Lachnidae. 
The  north-slope  colonies  showed  incomplete  life  cycles. 
All  insects  seen  there  were  virginoparae  apterae  or  alatae 
and  their  larvae.  Colonies  were  small  in  number  and 
were  developed  from  winged  virginoparae  from  other 
areas  that  colonized  plants  on  north-facing  slopes  from 
other  areas  (table  3). 

The  order  of  appearance  of  particular  forms  on 
Lucksalp  was  earlier  than  on  the  Stillberg  research  area. 
The  first  observation,  made  in  the  last  third  of  May, 
shows  68  percent  unhatched  eggs,  and  all  fundatrix  lar- 
vae were  part  of  the  first  two  larval  instars.  At  the  same 
time  on  Stillberg  87  to  90  percent  of  the  eggs  were 
unhatched. 

In  the  first  third  of  June  the  colony  structure  on 
Lucksalp  was:  7  percent  of  the  eggs  were  unhatched, 
young  fundatrices  in  all  four  larval  instars  were  found, 
but  no  adults.  A  similar,  but  slightly  different,  situation 
was  recorded  on  Stillberg.  There  were  25  to  38  percent 
unhatched  eggs,  and  the  colonies  contained  only  the  first 
three  larval  instars. 

During  the  observations  in  the  second  third  of  June  we 
found  adult  fundatrices  on  Lucksalp  and  on  the  east 
slopes  of  Stillberg.  With  the  appearance  of  fundatrices, 
the  density  of  populations  increased  gradually. 

On  the  other  Stillberg  slopes  (ENE  and  NE)  fundatrices 
were  found  10  days  later.  Simultaneously  we  recorded 
adults  of  apterous  virginoparae  on  Lucksalp.  This  is  a 
time  of  heavy  increases  in  population,  because  the  multi- 
plication rate  of  apterous  virginoparae  is  very  high,  and 
the  duration  of  larval  instars  very  short. 

In  an  experiment  that  was  conducted  under  conditions 
of  changeable  room  temperature,  one  apterous  virgino- 
parae was  isolated  on  a  twig  that  was  kept  moist.  Ten 
days  later,  we  recorded  six  apterous  virginoparae  of 
a  new  generation  (which  started  with  larviposition);  16 
in  the  fourth  larval  instar;  and  18  larvae  were  in  yoimger 
stadiums. 

It  was  presumed  that  the  effect  of  the  multiplication 
rate  is  not  the  same  in  nonisolated  field  conditions,  but 
on  some  trees  the  increments  of  population  for  10  days 
increased  by  a  factor  of  10  or  more.  On  the  other  hand, 
some  other  populations  were  substantially  reduced  by  the 
action  of  parasites  and  predators. 

In  the  first  third  of  July  at  Lucksalp  a  typical  popula- 
tion consisted  of  more  than  170  individuals;  almost  all 
belonged  to  apterous  virginoparae  larvae  (83  percent)  or 
adult  insects  ( 15  percent).  The  rest  (2  percent)  were 
alatoid  larvae  of  alatae  virginoparae.  The  fundatrix  gen- 
eration was  dead.  At  Stillberg,  populations  were  smaller 


with  most  of  the  population  made  up  of  larvae  and  adults 
of  apterous  virginoparae;  however,  fundatrices  were  also 
found  to  be  present.  The  fourth  larval  instar  of  alatae 
virginoparae  was  recorded  on  the  east  slopes  only.  The 
proportion  of  alatoid  larvae  was  10  percent  of  the  total 
number  of  the  fourth  larval  instar. 

The  first  Swiss  stone  pine  aphid  was  recorded  on  the 
north  slopes  of  Stillberg  in  the  middle  of  July.  Colonies 
were  small  and  consisted  of  only  a  few  adult  alatae 
virginoparae  and  young  larvae.  Few  of  the  colonies  were 
without  adults.  The  population  structures  of  the  other 
slopes  were  unchanged  in  content  until  the  last  third  of 
August.  Only  the  numbers  of  individuals  were  variable, 
depending  on  weather  conditions. 

The  rapid  decrease  in  number  of  individuals  at  the  be- 
ginning of  August  was  caused  by  low  temperatures  and 
snow.  The  greatest  changes  in  population  quantity  were 
recorded  on  north  slopes.  The  average  population  struc- 
ture in  the  course  of  July  and  August  was:  The  majority 
of  individuals  belonged  to  various  larval  instars,  and  the 
adult  part  of  the  population  consisted  of  apterae  and 
alatae  viviparous  parthenogenetic  females.  Average  ratio 
of  winged  and  wingless  forms  was  1:5.3  with  an  increas- 
ing tendency  to  winged  forms  in  colder  sites.  This  ten- 
dency is  due  to  the  strategy  of  the  species  to  avoid  expo- 
sure to  more  extreme  climatic  conditions  (by  colonizing 
another  host  plant  of  the  same  species,  but  in  better  envi- 
ronmental conditions).  Extremes  of  the  ratio  were  1:8  and 
1:2.  The  first  was  recorded  on  Lucksalp,  and  the  second 
on  Stillberg  north  slopes. 

The  highest  population  numbers  were  found  in  the  last 
third  of  July,  after  which  the  numbers  decreased  rapidly 
as  a  consequence  of  low  temperatures  at  the  beginning 
of  August.  Thereafter,  the  population  number  increased 
quickly  again  to  a  secondary  peak  at  the  end  of  August. 
Occurrence  of  the  last  generation  was  recorded  after  an- 
other cool  weather  period  at  the  end  of  August.  As  the 
overwintering  eggs  are  produced  sexually,  the  last  genera- 
tion of  the  year  must  contain  males  and  sexual  females, 
instead  of  the  parthenogenetic  forms  that  occur  during 
the  rest  of  the  year.  However,  sexuales  do  not  occur  un- 
der all  investigated  conditions.  On  the  north  slopes  of 
Stillberg,  for  example,  no  sexuales  forms  were  recorded. 
On  Lucksalp,  occurrence  of  sexuales  forms  was  about 
10  days  before  Stillberg. 

The  first  appearance  of  eggs  was  observed  at  the  begin- 
ning of  September.  Oviparous  females  move  away  from 
the  colony  to  the  top  part  of  current  shoots,  and  lay  over- 
wintering eggs  in  rows  on  needle  clusters.  Eggs  were 
brown  and  later  black  and  shiny  48  hours  after  being  laid. 
Average  number  of  eggs  in  the  row  along  one  needle  was 
8.2  and  the  extremes  were  16  and  1.  Maximum  egg  num- 
ber per  one  colony  was  316,  recorded  in  the  first  week  of 
September  at  Lucksalp.  The  same  colony  contained  eight 
sexuales  females.  Results  of  dissection  of  females  show 
12  ripe  eggs  in  the  ovariole  on  average.  This  was  smaller 
in  comparison  with  the  embryo  numbers  obtained  by  dis- 
section of  parthenogenetic  females  during  July,  which 
was  30. 


114 


Table  3 — Average  number  of  different  morphs  of  Cinara  cembrae  on  Swiss  stone  pine  sample  trees  from  last  third  of  May  to  second  third  of 
September 




Morph 

May 

June 

Julv 

August 

September 

3/3 

1/3 

2/3 

3/3 

1/3  2/3 

3/3 

1/3 

2/3 

3/3 

1/3 

2/3 

Lucksalp  (SW  slope) 

Eqqs 

5.6 

0.6 

60.0 

93.0 

Larvae 

2.6 

7.6 

35.0 

103.0 

148.0  178.0 

199.0 

68.0 

82.0 

124.0 

9.0 

11.0 

Fundatrices 

1.2 

1.5 

Apterae  virginoparae 

2.0 

27.0  12.0 

24.0 

8.3 

18.0 

24.0 

7.3 

6.7 

Aiatae  virginoparae 

1.5 

2.4 

1.8 

1.5 

2.0 

0.2 

Oviparous  females 

0.1 

4.9 

5.4 

Apterae  males 

0.1 

0.3 

1.4 

Aiatae  males 

0.1 

Stillberg  (E  slope) 

Eqqs 

7.2 

2.0 

0.2 

3.8 

57.0 

Larvae 

0.8 

4.4 

30.0 

41.0 

58.0  100.0 

110.0 

54.0 

80.0 

104.0 

5.9 

8.1 

Fundatrices 

0.6 

1.8 

1.0 

Apterae  virginoparae 

3.6  30.0 

31.0 

6.3 

11.0 

20.0 

2.9 

3.2 

Aiatae  virginoparae 

3.0 

5.0 

5.8 

0.7 

Oviparous  females 

0.2 

2.3 

Apterae  males 

0.1 

0.3 

Aiatae  males 

0.1 

0.2 

Stillberg  (ENE  slope) 

Eqqs 

7.0 

3.0 

1.1 

Larvae 

0.8 

4.8 

6.7 

34.0 

124.0  140.0 

145.0 

54.0 

69.0 

93.0 

4.6 

8.3 

Fundatrices 

2.4 

0.2 

Apterae  virginoparae 

5.0  12.0 

18.0 

3.2 

10.0 

17.0 

3.0 

3.5 

Aiatae  virginoparae 

2.6 

1.0 

3.8 

4.7 

1.8 

Oviparous  females 

1.4 

5.2 

Apterae  males 

0.8 

Aiatae  males 

0.2 

0.5 

Stillberg  (NE  slope) 

Eggs 

6.5 

2.0 

8.0 

Larvae 

1.0 

5.9 

7.7 

32.0 

40.0  89.0 

118.0 

26.0 

40.0 

58.0 

8.6 

12.0 

Fundatrices 

2.3 

0.8 

Apterae  virginoparae 

1.8  10.0 

18.0 

2.3 

7.1 

9.4 

3.8 

9.2 

Aiatae  virginoparae 

0.9 

1.0 

1.3 

3.0 

2.6 

2.8 

Oviparous  females 

0.8 

2.0 

Apterae  males 

0.7 

1.5 

Aiatae  males 

Stillberg  (N  slope) 

Eggs 

Larvae 

5.8 

43.0 

1.0 

11.0 

18.0 

Fundatrices 

Apterae  virginoparae 

6.2 

1.0 

2.0 

4.8 

Aiatae  virginoparae 

1.0 

3.4 

1.6 

1.2 

Oviparous  females 
Apterae  males 
Aiatae  males 


115 


SEASONAL  MOVEMENTS 

The  most  important  factors  affecting  aphid  feeding  and 
nutrition  are  the  change,  in  time,  of  growth  and  develop- 
ment of  the  host  plant  (Klungauf  1987).  In  conjmiction 
with  this  change,  seasonal  movements  of  colonies  were 
observed. 

The  fundatrix  generation  moves  from  the  needles  where 
overwintering  eggs  were  laid.  Feeding  sites  of  fimda- 
trices,  and  later  generations  of  virginoparae,  in  the  course 
of  Jime  and  first  half  of  July  were  on  last  year's  shoots. 
During  July,  colonies  moved  to  twigs  and  branches  in 
lower  parts  of  the  crown  (in  extreme  cases  some  of  the 
colonies  were  fovmd  on  stems).  This  was  just  after  the 
peak  of  the  annual  increment  of  the  shoots,  when  they 
start  to  become  lignified.  It  seems  to  be  related  to  de- 
scending sap  movement.  Conifer  lachnids,  as  phloem 
feeders,  exploit  food  sources  in  lower  parts  of  the  plant 
that  become  available  as  the  season  progresses. 

The  alatoid  larvae  of  alatae  virginoparae  usually  move 
from  the  colony  to  the  tops  of  the  shoots  before  their  final 
moult.  Here  they  moult  into  alatae  adults  and  fly  off  to 
colonize  similar  positions  on  other  plants. 

Simultaneously  with  the  occurrence  of  sexuales,  the 
colonies  moved  ascendentally.  At  the  end  of  the  season 
the  majority  of  the  individuals  were  on  last  year's  and 
current  season's  shoots.  From  this  position  oviparous 
females  moved  to  lay  eggs  on  needles. 

Additional  trials  were  carried  out  with  colonies  isolated 
by  polytene  covers,  which  prevented  migration  from  top  to 
bottom  parts  of  the  host  plant.  In  the  beginning,  the  den- 
sity of  isolated  colonies  was  much  higher  than  the  density 
of  nearby  nonisolated  colonies.  However,  early  occurrence 
of  alatae  forms  indicated  an  attempt  to  change  the  colo- 
nies' position. 

Later,  complete  starvation  of  the  population  was  caused 
by  restriction  of  movement.  It  was  concluded  that  the 
life  cycle  of  Swiss  stone  pine  aphid  is  very  much  synchro- 
nized with  the  growth  and  development  characteristics  of 
their  host  plant,  and  in  conjunction  with  this,  quality  of 
nutrition. 

PEST  DAMAGES 

Keen  (1938)  reported  that  several  species  of  lachnids 
cause  considerable  injury  to  conifers  in  North  America. 
Fumiss  and  Carolin  (1977)  described  visible  signs  of  in- 
jury, such  as  yellowing  of  needles  or  reduction  of  growth 
of  young  trees.  According  to  Seithner  (1936),  the  damage 
to  Swiss  stone  pine  is  quite  iinimportant,  but  he  believed 
that  the  trees  could  be  weakened  by  the  sucking  and  that 
secondary  pests  (for  example,  Pissodes  and  Pityophthorus) 
could  easily  attack  them. 

We  found  no  direct  visible  signs  of  damage  by  Swiss 
stone  pine  aphid  except  for  some  discolorations  on  feeding 
sites.  Our  observations  show  that  retardation  of  the 
growth  increment  by  Swiss  stone  pine  aphid,  if  any,  is 
much  less  significant  than  retardation  due  to  other  envi- 
ronmental factors. 

Indirectly,  Swiss  stone  pine  aphid's  honeydew  excre- 
tion, in  the  absence  of  heavy  rain,  accumulates  on  the 
surface  of  the  shoots  and  forms  an  ideal  substrate  for 


saprophytic  sooty  mould  fungi  to  develop.  A  black  cover- 
ing of  hyphae  and  spores  can  reduce  the  quality  of  trees 
that  are  grown  in  parks,  but  for  the  Stillberg  and 
Lucksalp  research  areas  it  is  not  so  important.  On  the 
other  hand,  honey  derived  from  honeydew,  particularly 
from  conifer  lachnids,  is  highly  valued  and  of  considerable 
economic  significance.  Fossel  (1971)  found  that  the  Swiss 
stone  pine  aphid  produced  copious  honeydew  in  the  movm- 
tainous  districts  of  Austria. 

CONCLUSIONS 

Information  gathered  during  this  study  has  allowed  the 
following  conclusions  to  be  drawn: 

•  Swiss  stone  pine  aphid  was  the  most  common  insect 
pest  on  Swiss  stone  pine  used  for  afforestation  in  the  re- 
search areas  of  the  Dischma  Valley. 

•  The  morphological  characteristics  of  specimens  and 
the  chronological  series  of  morphs  found  on  Stillberg  and 
Lucksalp  are  similar  to  the  conventional  descriptions. 
However,  for  some  site  aspects  of  the  Dischma  Valley  the 
life  cycle  of  Swiss  stone  pine  aphid  was  incomplete.  The 
life  cycle  abnormality  was  observed  on  the  north  slopes  of 
Stillberg. 

•  The  order  of  appearance  of  particular  insect  forms, 
structure,  and  number  of  individuals  in  colonies  also  were 
strongly  related  to  site  aspects.  This  is  particiilarly  true 
for  the  temperature  effect  that  caused  later  development 
and  smaller  sizes  of  colonies  on  the  north  slopes. 

•  Swiss  stone  pine  aphid  was  observed  to  cause  only 
unimportant  discolorations  on  the  feeding  sites.  The 
damage  to  the  trees  is  much  less  significant  than  that  pro- 
duced by  the  normally  harsh  environmental  conditions  of 
the  upper  timberline  area. 

When  looked  at  in  isolation,  the  damage  produced  by 
the  pest  observed  in  this  study  cannot  be  considered  seri- 
ous. However,  it  is  only  one  link  in  a  chain  of  factors  det- 
rimental to  the  growth  of  afforestation  plants.  How  the 
insect  interacts  with  other  factors  harmful  to  tree  growth 
(for  example,  fungi)  is  as  yet  unknown,  and  therefore  the 
damage  caused  by  colonies  of  Swiss  stone  pine  aphid  can- 
not be  accurately  assessed. 

Until  it  is  known  how  detrimental  the  aphid  is,  the  im- 
portance of  control  measures  or  predators  and  parasites 
also  cannot  be  estimated. 

Clearly,  6  months  investigation  is  too  short  for  con- 
clusive results.  However,  the  information  gained  has 
provided  an  important  first  step  in  imderstanding  the 
relationship  between  the  insect  pest,  Swiss  stone  pine 
aphid,  and  the  host  plant,  Swiss  stone  pine,  in  high- 
altitude  afforestation. 

REFERENCES 

Alikhan,  M.  A.  1960.  The  experimental  study  of  chemotac- 
tic  basis  of  host  specificity  in  a  phytophagous  insect. 
Aphis  fabae  Scop.  (Aphididae:  Homoptera).  Annales 
Univ.  Mariae  Curie  -  Sklodowska,  Sectio  C.  15:  117-158. 

Carter,  C.  I.;  Maslen,  N.  R.  1982.  Conifer  lachnids  in 
Britain.  Alice  Holt  Lodge,  Farnham,  Surrey,  Engleind: 
Forestry  Commission  Research  Station. 


116 


Chapman,  R.  F.;  Bemays,  E.  A.;  Simpson,  S.  J.  1981. 
Attraction  and  repulsion  of  the  aphid,  Cavariella 
aegopodii,  by  plant  odors.  Journal  of  Chemical  Ecology. 
7:  881-888. 

Dixon,  A.  F.  G.  1987.  Parthenogenetic  reproduction  and 
the  rate  of  increase  in  aphids.  In:  Aphids:  their  biology, 
natural  enemies,  and  control.  2A:  269-287. 

Eastop,  V.  F.  1972.  A  taxonomic  review  of  the  species 
of  Cinara  Curtis  occurring  in  Britain  (Hemiptera: 
Aphididae).  Bulletin  of  the  British  Museimi,  Natiiral 
History  (Entomology).  27:  103-186. 

Fossel,  A.  1971.  New  observation  on  Cinara  cembrae 
(Homoptera,  Lachnidae).  Annales  Zoologici.  28:  353-365. 

Fumiss,  R.  L.;  Carolin,  V.  M.  1977.  Western  forest  in- 
sects. Misc.  Publ.  1339.  Washington,  DC:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service.  654  p. 

Hille  Ris  Lambers,  D.  1955.  Hemiptera  2.  Aphididae. 
Zoology  of  Iceland.  3(52a):  1-29. 

Keen,  F.  P.  1938.  Insect  enemies  of  western  forests.  Misc. 
Publ.  273.  Washington,  DC:  U.S.  Department  of 
Agriculture. 

Klingauf,  F.  A.  1987.  Feeding,  adaptation  and  excretion. 
In:  Aphids:  their  biology,  natural  enemies,  and  control. 
2A:  225-253. 


Lees,  A.  D.  1961.  Clonal  polymorphism  in  aphids.  In:  In- 
sect polymorphism.  Sjmiposia  of  the  Royal  Entomologi- 
cal Society  of  London.  1:  68-79. 

Pintera,  A.  1966.  Revision  of  the  genus  Cinara  Curt. 
(Aphidodea,  Lachnidae)  in  Middle  Europe.  Acta  ent. 
bohemoslov.  63:  281-321. 

Pospisil,  J.  1976.  Sensory,  olfactory  and  visual  orientation 
of  insects  of  the  tropical  region  III.  Visual  and  olfactory 
orientation  of  alate  forms  of  the  aphid,  Aphis  gossypii 
Glover,  in  Cuba.  Academia  de  Ciencias  de  Cuba,  Serie 
Biologica.  66:  17-23. 

Schonenberger,  W.  1984.  Performance  of  a  high  altitude 
afforestation  under  various  site  conditions.  Proc.  3d 
lUFRO  Workshop:  233-240. 

Seither,  M.  1936.  Lachnus  cembrae  an.  sp.,  die  Zirben- 
blattlaus.  Centralbl.  ges.  Forstwes.  62:  33-49. 

Stroyan,  H.  L.  G.  1955.  Recent  additions  to  the  British 
aphid  faima.  Part  II.  Transactions  of  the  Royal  Entomo- 
logical Society  of  London.  106:  283-340. 

Stroyan,  H.  L.  G.  1960.  Three  new  subspecies  of  aphids 
from  Iceland  (Hem.,  Hom.).  Entomologiske  Meddelelser. 
29:  250-265. 


117 


GENETIC  CONSEQUENCES  AND 
RESEARCH  CHALLENGES  OF  BLISTER 
RUST  IN  WHITEBARK  PINE  FORESTS 

Raymond  J.  Ho^ 
Susan  K.  Hagle 
Richard  G.  Krebill 


Abstract — Susceptibility  of  whitebark  pine  {Pinus  albicaulis)  to 
blister  rust  (caused  by  Cronartium  ribicola)  is  reviewed.  Prog- 
ress is  reported  on  studies  that  assess  the  level  of  susceptibility 
over  its  entire  range  and  the  existence  of  resistance  in  various 
stands.  Two  breeding  approaches  are  discussed:  (1)  the  tradi- 
tional, where  trees  are  selected,  tested,  then  established  in  seed 
orchards;  (2)  a  natural  approach  that  aids  natural  processes  to 
establish  "a  natural  selection  stand." 


In  1910,  eastern  white  pine  (Pinus  strobus)  seedlings 
that  had  been  grown  in  Europe  were  planted  in  the  area 
of  Point  Grey,  BC,  Canada.  Some  of  these  seedlings  were 
infected  with  white  pine  blister  rust  (caused  by  the  fungus 
Cronartium  ribicola).  The  disease  was  finally  noticed  in 
the  fall  of  1921,  when  it  was  observed  on  endemic  western 
white  pine  (Pinus  monticola).  The  destructive  nature  of 
this  disease  in  North  American  white  pines  had  already 
been  documented  in  Europe  (Spaulding  1911). 

What  followed  was  a  frantic  rush  to  stop  blister  rust's 
spread  by  destroying  infected  trees,  eradicating  currants 
and  gooseberries  (genus  Ribes)  that  are  alternate  hosts  for 
the  disease,  and  using  several  promising  fungicides.  But 
this  work  failed,  and  the  fungus  kept  spreading.  By  about 
1960,  the  fungus  had  spread  throughout  most  the  range 
of  whitebark  pine  (Pinus  albicaulis)  (fig.  1),  which  also  in- 
cludes most  of  the  range  of  western  white  pine  and  sugar 
pine  (Pinus  lambertiana).  Examples  of  blister  rust  on 
whitebark  pine  are  illustrated  in  figure  2. 

The  first  infected  whitebark  pine  recorded  was  in 
the  arboretum  of  the  University  of  British  Columbia, 
Vancouver,  BC,  in  1922  (Bedwell  and  Childs  1943).  The 
first  discovery  of  blister  rust  on  native  whitebark  pine 
was  in  1926  on  the  Birkenhead  River  in  the  Coast  Range 
of  British  Columbia,  Canada  (Childs  and  others  1938). 
Intense  infection  and  mortality  by  blister  rust  was  not  far 
behind  (Bedwell  and  Childs  1943). 

The  degree  of  infection  was  100  percent  in  many  stands 
in  the  northern  areas,  and  decreased  to  the  south  where 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Raymond  J.  HofF  is  Research  Plant  Geneticist,  Intermountain  Research 
Station,  Forest  Service,  U.S.  Department  of  Agriculture,  Moscow,  ID  83843, 
USA.  Susan  K.  Hagle  is  Pathologist,  Northern  Region,  Forest  Service,  U.S. 
Department  of  Agriculture,  Missoula,  MT  59807,  USA.  Richard  G.  Krebill 
is  Assistant  Station  Director  and  Pathologist,  Intermountain  Research 
Station,  Forest  Service,  U.S.  Department  of  Agriculture,  Ogden,  UT  84401, 
U.S.A. 


whitebark  pine  occurs  in  higher  and  drier  sites.  White- 
bark pines  grovnng  below  about  45°  N.  lat.  in  Idaho  and 
Montana  have  much  less  rust  than  whitebark  of  higher 
latitudes.  However,  in  the  Cascade-Sierra  Nevada  Moun- 
tains the  rust  has  spread  to  about  lat.  36°  15'  N.,  where  it 


Pinus  albicaulis 

BH  malor  subalpine  component 


Figure  1 — Range  map  of  whitebark  pine  showing 
areas  of  high,  moderate,  and  little  or  no  blister  rust 
incidence. 


118 


has  been  especially  devastating  to  western  white  pine 
and  sugar  pine.  Whitebark  pine  too  has  been  heavily 
impacted  in  most  of  these  mountain  ranges.  But  no 
infection  was  found  on  whitebark  pine  at  Tioga  Pass, 
Yosemite  National  Park,  at  about  39°  N.  lat.  in  1992. 

The  relative  susceptibility  of  the  white  pines  to  blister 
rust  has  been  tested  by  several  workers.  Bingham 
(1972b)  siimmarized  these  reports  (table  1).  All  reports 
agreed  that  whitebark  pine  was  most  susceptible  of  the 
tested  five-needled  pines.  Fiu-ther  work  was  done  by 
Bedwell  and  Childs  (1943)  to  compare  the  relative  suscep- 
tibility of  whitebark  versus  western  white  pine.  These 
two  pines  grow  adjacent  for  much  of  their  ranges,  with 
whitebark  pine  at  the  high-elevation  sites  and  western 
white  pine  lower.  They  often  are  foimd  on  the  same  site 
in  the  overlap  zone.  Childs  and  Bedwell  (1948)  found  that 
susceptibility  (in  terms  of  nvunbers  of  cankers  and  speed 
of  tree  mortality)  of  whitebark  pine  was  several  times 
that  of  western  white  pine.  They  concluded  that  this  dif- 
ference was  due  in  part  to  (1)  a  higher  susceptibility  of 
current  year  whitebark  pine  needles  and  (2)  longer  needle 
retention  of  whitebark  pine  (5.3  years  for  whitebark  and 
3.8  years  for  western  white  pine). 


Table  1 — Comparison  of  susceptibility  in  white  pines  to  blister  rust 
by  Bingham  1972a  or  b  and  a  reshuffle  by  Hoff  and  others 
1980 


Species         Bingham  1972a  or  b      Hoff  and  others  1980 


Pinus  armandii 

1 

4 

P.  cembra 

1 

3 

P.  aristata 

1 

P.  wallichiana 

2 

6 

P.  koraiensis 

2 

1 

P.  peuce 

3 

5 

P.  sibirica 

4 

2 

P.  parviflora 

5 

3 

P.  strobiformis 

6 

8 

P.  strobus 

7 

11 

P.  flexilis 

8 

10 

P.  monticola 

9 

4 

P.  lambertiana 

10 

9 

P.  albicaulis 

11 

7 

119 


Bingham  (1983)  found  one  western  white  pine  in  10,000 
that  was  free  of  bhster  rust.  The  possibihty  of  finding 
disease-free  whitebark  pine  trees  seemed  pretty  sHm. 

Nonetheless,  at  sites  in  northern  Idaho  and  western 
Montana  where  mortahty  is  over  90  percent  due  to  blister 
rust,  we  have  found  a  few  individuals  with  no  or  just  a 
few  cankers.  Though  they  certainly  are  rare  (fig.  3).  This 
gives  us  hope  that  there  are  resistant  trees  that  can  be 
used  to  reestablish  populations  in  these  areas. 

Hofif  and  others  (1980)  presented  data  indicating  that 
three  phenotypically  resistant  whitebark  pine  trees  in- 
cluded in  a  species  trial  did  indeed  have  resistance  to  blis- 
ter rust.  In  this  same  study,  resistant  collections  of  west- 
em  white  pine  and  sugar  pine  were  also  included.  This 
resulted  in  a  reshuffle  of  the  white  pines  for  susceptibility 
to  blister  rust  (table  1).  In  this  comparison,  whitebark 
pine  moved  up  along  with  resistant  western  white  pine 
and  resistant  sugar  pine.  However,  these  same  species 
were  tested  in  France  by  Delatour  and  Birot  (1982),  in 
Japan  by  Yokota  (1983),  and  in  Germany  by  Stephan 
(1985),  and  the  resistance  level  of  the  collections  of  white- 
bark, western  white,  and  sugar  pine  was  much  lower 
(tables  2,  3,  4).  The  only  explanation  is  that  there  are 
different  races  of  blister  rust  at  these  other  test  sites. 

In  spite  of  this  bad  news,  our  plans  are  to  proceed  to  de- 
velop resistance  in  whitebark  pine,  including  as  many 


Figure  3 — Blister  rust-resistant  candidate  trees  at  Gisborne  Mountain  (A) 

120 


genes  for  resistance  as  possible.  If  new  races  of  the  fungus 
appear,  the  new  resistant  varieties  should  have  adequate 
flexibility  to  exhibit  resistance  to  the  new  races.  This 
seems  justified,  since  high  levels  of  resistance  in  western 
white  pine  and  sugar  pine  are  still  being  observed  in  blister 
rust  resistance  tests,  even  though  several  races  of  the  fun- 
gus are  evident. 

ACTIVE  WORK  AND  PRELIMINARY 
PROGRESS 

Blister  Rust  Surveys  of  Stands,  Clearcuts,  and 
Bums — Onr  objectives  are  to  determine  the  level  of  blis- 
ter rust  damage  over  the  reuige  of  whitebark  pine  and  to 
locate  phenotypically  resistant  trees.  As  mentioned,  the 
degree  of  damage  is  associated  with  latitude.  Phenotypi- 
cally resistant  trees  have  been  located  in  most  stands 
where  high  mortality  to  blister  rust  has  occurred.  We 
also  observed  a  relatively  large  amount  of  natural  regen- 
eration occurring  in  clearcuts  and  bums  adjacent  to 
stands  of  whitebark  pine  that  have  had  high  rust-caused 
mortality.  Table  5  shows  the  number  of  trees  per  acre 
and  the  degree  of  infection  by  blister  rust  in  four  such 
stands.  Natural  regeneration  of  whitebark  pine  occurs 
from  unused  seed  caches  of  Clark's  nutcracker.  The 


Burl^e's  Peak  (B). 


Table  2 — Percent  of  rust-infected  seedlings  in  bulk  collections  of 
white  pine  species  (from  Delatour  and  Birot  1 982) 


Geographic 

Number 

Percent 

group 

Species 

of  lots 

infected 

European 

P.  cembra 

12 

0 

P.  peuce 

7 

0-20 

Asian 

P.  sibirica 

2 

0 

P.  parviflora 

3 

0 

P.  koraiensis 

7 

0-9 

P.  wallichiana 

10 

7-67 

P.  armandii 

1 

58 

American 

r.  ailolala 

1 

o 
c. 

P.  albicaulis 

5 

40-100 

P.  flexilis 

17 

0-100 

P.  strobiformis 

9 

40-100 

P.  lambertiana 

5 

64-96 

P.  strobus 

24 

88-100 

P.  monticola 

5 

94-100 

abundance  of  regeneration  was  surprising.  It  immedi- 
ately broxight  up  questions:  At  what  level  of  mortality  will 
there  not  be  enough  seed  available  for  caching  by  the  nut- 
cracker? If  the  seed  did  not  come  from  the  adjacent  high- 
mortality  stand,  where  did  it  come  from? 

Artificial  Inoculation  With  Blister  Rust — The  objec- 
tive is  to  determine  the  level  and  variety  of  mechanisms 
that  impart  resistance  in  phenotypically  resistant  trees 
and  in  stands  that  have  had  high  mortality  by  blister 

Table  3 — Results  of  inoculation  in  France  and  Idaho. 

Percent  of  rust  infection  on  "resistant"  crosses 
of  P.  lambertiana  and  P.  monticola  from 
inoculation  in  France  or  Idaho  (Delatour  and 
Birot  1 982) 

Seed  lot  No.  France 
and  cross  mean  Idaho 


 Percent 

P.  lambertiana 


41  CF^ 

99 

55 

41  BF, 

98 

57 

41  AF, 

52 

25 

P.  monticola 

43  BF, 

96 

40 

43  EF, 

93 

30 

43  CF, 

76 

40 

43  A  F, 

32 

30 

43  FF, 

73 

65 

43  DF, 

35 

45 

47  A  F, 

75 

65 

46HF2 

63 

35 

46EF2 

56 

40 

46BF2 

55 

45 

46FF2 

44 

35 

46G  F2 

40 

30 

46DF2 

38 

20 

46  AFj 

37 

30 

46CF2 

24 

15 

Table  4 — Comparison  of  blister  rust  susceptibility  of  some  white 
pines  from  Stephan  (1985) 


North  America 

Percent 

Eurasia 

Percent 

species 

infection 

species 

infection 

P.  albicaulis 

97 

P.  armandii 

0 

P.  aristata 

66 

P.  cembra 

0 

P.  balfouriana 

90 

P.  koraiensis 

23 

P.  flexilis 

98 

P.  morrisonicola 

40 

P.  lambertiana 

^97  (76) 

P.  parviflora 

22 

P.  monticola 

99 

P.  peuce 

22 

P.  strobiformis 

88 

P.  pumila 

0 

P.  strobus 

100 

P.  sibirica 

17 

P.  wallichiana 

40 

'Two  seedlots  were  tested;  one  was  not  as  susceptible  as  the  other. 


rust.  A  trial  test  was  inoculated  in  August  1991  using 
methods  outlined  by  Bingham  (1972a).  This  test  included 
about  1,000  seedlings  from  36  families  from  10  populations. 
The  inoculation  was  light;  only  26  percent  of  the  seedlings 
developed  needle  spots.  However,  there  was  considerable 
variation  among  populations  (table  6).  Even  with  the  light 
inoculation,  it  was  encouraging  to  observe  several  seedlings 
that  had  what  we  call  "premature  needle  shed"  (McDonald 
and  Hoff  1970).  This  is  a  resistance  mechanism  in  which 
needles  with  needle  spots  are  shed  from  9  to  12  months 
after  inoculation  and  before  the  rust  fungus  has  invaded 
the  stem.  This  type  of  resistance  has  been  observed  in 
several  white  pine  species  (Hoff  and  others  1980).  The 
seedlings  were  reinoculated  in  August  1992  and  as  of 
November  1,  1992,  87  percent  had  needle  spots. 

Adaptive  Variation  of  Whitebark  Pine — How  far 

can  we  transfer  seed  £ind  not  cause  maladaptation  in  the 
seedlings?  The  range  of  whitebark  pine  extends  nearly 
19  degrees  of  latitude  and  20  degrees  of  longitude,  and 
from  1,524  to  3,354  m.  In  addition,  the  species  grows  in 
various  ecological  situations.  It  can  fill  the  role  of  a  pio- 
neer and  a  serai.  It  can  exist  for  long  periods  as  a  domi- 
nant on  harsh  sites  and  even  in  some  stands  with  the  cli- 
max species,  subalpine  fir.  Whitebark  pine  is  found  in 
106  of  225  habitat/phase  types  in  eastern  Idaho-western 
Wyoming,  central  Idaho,  and  Montana  (Pfister  and  others 
1977;  Steele  and  others  1981,  1983). 

However,  whitebark  pine  is  relatively  intolerant  to 
shade  and  highly  sensitive  to  competition  (Arno  and 
Weaver  1990).  Therefore  it  is  frequently  restricted  to 
tougher  sites  that  prevent  or  restrict  the  growth  of  other 


Table  5 — Number  and  infection  of  whitebark  pine  seedlings  in  four 
clearcuts 


Name  of  Seedlings  Blister  rust 

clearcut  per  hectare  infection 


Percent 

Vermillion  Pass 

445 

55 

Upper  Coal  Creek 

1,544 

4 

Upper  Big  Creek 

884 

33 

Divide  Mountain 

3,929 

52 

121 


Table  6 — Variation  in  various  traits  of  whitebark  pine  for  1 0  provenances 


Preliminary  population  comparisons 


Third-year 

riea 

hXi 

blister 

^Ji*  Aif  An  Q  n  A 

1  a* 

Lai. 

Long. 

Elevation 

height 

stem 

DUO 

rust 

m 

cm 

—  Percent-  - 

Cooper  Pass 

47-32N 

115-44W 

1,494 

17.1 

22 

9 

22 

Gisborne 

48-21 N 

116-44W 

1,692 

15.0 

10 

4 

24 

Freezeout 

47-01 N 

116-02W 

1,707 

10.6 

3 

0 

5 

Lunch  Peak 

48-22N 

116-22W 

1,982 

16.9 

10 

9 

33 

Brundage  Mountain 

45-01 N 

116-07W 

2,195 

16.2 

17 

8 

24 

Seven  Devils 

45-21 N 

11 6-31 W 

2,302 

12.8 

6 

3 

26 

Saddle  Mountain 

45-42N 

113-59W 

2,378 

18.1 

24 

2 

46 

Porphyry  Peak 

46-49N 

110-44W 

2,509 

12.0 

0 

13 

33 

Boulder  Peak 

41-35N 

1 23-05 W 

2,523 

10.7 

0 

0 

Palmer  Mountain 

45-04N 

110-35W 

2,652 

16.3 

0 

22 

41 

Mean 

14.6 

9 

7 

25 

species.  Because  of  the  combination  of  several  severe  en- 
vironmental factors,  such  as  short  growing  season,  cold 
air,  and  snow  blast  (McCaughey  and  Schmidt  1990), 
whitebark  pine  forms  krummholz  stands  of  shrublike 
trees  at  or  near  timberline.  At  lower  elevations,  where 
growing  conditions  are  not  quite  so  severe,  whitebark  pine 
grows  in  nearly  pure  stands  of  trees  30  to  90  feet  tall 
(McCaughey  and  Schmidt  1990).  When  environmental 
conditions  are  even  less  severe,  whitebark  pine  is  associ- 
ated with  other  tree  species  and  takes  on  a  form  much 
like  lodgepole  pine.  When  growdng  conditions  are  ideal, 
whitebark  pine  seedlings  will  grow  fairly  rapidly  (fig.  4). 

With  such  a  wide-ranging  and  ecologically  diverse  spe- 
cies one  would  expect  several  geographic  races.  Since 
nutcrackers  normally  cache  the  seed  at  least  several  him- 
dred  meters  and  in  many  cases  several  kilometers  from 
its  source,  it  is  also  easy  to  specvilate  that  there  are  few 
races  because  genes  from  several  diverse  populations  are 
being  continually  mixed.  Future  tests  should  clarify  this 
puzzle. 

Seed  from  about  40  stands  have  been  collected.  Our 
first  area  of  study  will  be  Idaho  and  Montana.  The  test 
will  be  established  with  seed  from  about  100  populations. 
A  preliminary  test  has  revealed  some  interesting  data 
(table  6).  There  is  considerable  variation  among  the  popu- 
lations in  3-year-old  total  height  (fig.  4).  The  tallest  seed- 
lings were  from  the  Saddle  Mountain  stand — a  population 
from  high  elevation.  There  is  also  high  variation  among 
families  within  stands  (table  7).  Also,  characteristics  such 
as  red  coloration  of  the  succulent  stem  and  the  prolifera- 
tion of  lateral  buds  vary  with  population  (fig.  5).  So  far, 
none  of  this  variation  can  be  associated  with  geographic 
position  or  elevation. 

Inbreeding — In  many  stands,  only  a  few  individuals 
remain.  This  results  in  much  higher  opportimity  for  self- 
fertilization  (inbreeding).  At  what  level  will  inbreeding 
adversely  affect  growth  and  survival  of  the  seedlings? 
A  study  has  been  initiated  to  compare  growth  and  other 
traits  between  a  highly  decimated  population  and  a  stand 
with  very  low  mortality.  Trees  at  the  two  sites  will  be  in- 
tercrossed, crossed  with  pollen  from  the  same  site,  and 
compared  v^dth  wind-pollinated  seed. 


BREEDING  PLANS  TO  DEVELOP 
RESISTANCE 

Our  surveys  have  shown  that  there  are  many  pheno- 
typically  resistant  whitebark  pine  trees.  A  traditional 
breeding  approach  could  be  used  to  develop  a  new  blister 
rust-resistant  variety  of  whitebark  pine  from  these  trees. 
Three  options  in  this  approach  will  be  discussed.  How- 
ever, because  of  the  high  level  of  natural  regeneration  of 
whitebark  pine  in  clearcuts  and  burns  adjacent  to  high- 
mortality  stands,  another  approach  appears  more  attrac- 
tive. This  is  a  more  natural  method  using  mass  selection 
as  the  genetic  selection  system,  but  within  a  natural  set- 
ting and  with  natural  selection  processes. 

The  Traditional  Approach  (fig.  6) — Steps  to  success 
are: 

1.  Locate  blister  rust-resistant  phenotypes  of  white- 
bark pine.  Most  stands  have  a  few  candidate  trees,  but 
a  high  number  is  required  so  that  we  can  be  assured  of 


Figure  4 — Three-year-old  container-grown  white- 
bark pine. 


122 


Table  7 — Variation  among  whitebark  pine  families  within  the 
Gisborne  Stand 


'amily 

Third-year 
neignt 

Red  stem 

Prolific 

DUO 

Diister 
rust 

cm 

-  -  Percent  — 

1 

18 

6 

0 

11 

2 

12 

11 

0 

22 

3 

16 

0 

0 

29 

4 

14 

12 

0 

0 

5 

15 

20 

5 

22 

6 

14 

7 

0 

14 

7 

17 

0 

0 

7 

8 

15 

4 

13 

33 

9 

18 

20 

4 

65 

10 

15 

9 

0 

35 

11 

12 

5 

9 

30 

Mean 

15 

9 

3 

24 

finding  several  genes  for  resistance.  Thus,  many  stands 
will  have  to  be  visited.  The  breeding  population  for  west- 
em  white  pine  is  over  3,000  phenotypically  resistant  trees 
and  is  considered  a  minimum  nimiber  of  parent  trees  to 
assure  maintenance  of  resistance  against  several  races  of 
the  rust  fungus. 

2.  Collect  wind-pollinated  seed  from  each  tree.  In  spe- 
cial cases,  where  there  are  just  one  or  two  trees,  artificial 
pollination  by  other  candidate  trees  may  be  advisable. 

3.  Sow  seed,  grow  seedlings,  and  inoculate  with  blister 
rust  in  the  fall  of  their  second  growth  period. 

4.  Data  over  the  next  4  years,  after  inoculation,  will 
reveal  the  most  resistant  parents  of  the  phenotypically 
resistant  candidate  trees,  along  with  their  resistant 
progeny. 

Options  for  developing  the  new  resistant  variety  are: 

1.  Selected  resistant  parents  can  be  grafted  into  a  seed 
orchard. 

2.  Resistant  seedlings  can  be  used  to  establish  a  seed 
orchard. 

3.  Surviving  seedlings  from  rust  tests  that  were  not 
used  in  the  seed  orchard  can  be  outplanted  in  a  natural 
site.  The  numbers  for  this  planting  can  probably  be  in- 
creased by  rooted  cuttings  of  the  surviving  seedlings. 

Timing:  The  first  five  steps  cotdd  be  done  in  about  10 
years.  Time  to  flowering  of  grafts  could  be  10  years,  but  it 
will  probably  take  from  40  to  50  years  for  their  progeny  to 
flower.  New  technology  may  shorten  this  time. 

Problems: 

1.  What  is  the  natural  level  of  resistance? 

2.  How  far  can  seed  be  transferred  without  serious 
maladaptation? 

Natural  Selection  Stand — In  many  localities  in  Idaho 
and  Montana  whitebark  pine  has  been  decimated  by  blis- 
ter rust,  leaving  just  a  few  individuals  per  hectare.  Many 
of  these  remnant  stands  are  adjacent  to  burns  or  clear- 
cuts.  In  at  least  several  dozen  cases  these  openings 


contain  high  nvunbers  of  whitebark  pine  seedlings  and 
saplings  (table  5).  With  careful  management  we  may  be 
able  to  use  these  newly  established  stands  for  our  piir- 
pose.  Only  three  major  steps  need  to  be  taken: 

1.  Select  the  areas  within  the  clearcuts  and  bxarns  that 
will  be  managed  for  whitebark  pine  (fig.  7A). 

2.  Clean  the  selected  area  of  competing  trees  and 
shrubs  (fig.  7B).  Whitebark  pine  is  a  poor  competitor.  In 
fact  it  probably  can  be  grown  at  much  lower  elevations  if 
the  site  is  kept  free  of  competition. 

3.  Let  nature  take  its  course  concerning  the  selection  of 
the  most  resistant  trees  (figs.  7C,  D). 

Problems: 

1.  What  is  the  natural  level  of  resistance?  We  need  to 
know  the  level  of  resistance  (1,  5,  or  10  percent)  in  stands 
to  choose  sites  that  have  a  good  chance  for  success.  If  only 
1  percent  of  the  seedlings  are  resistant  when  produced  by 
a  mature  stand  that  has  had  90  to  95  percent  mortality 
by  blister  rust,  a  candidate  area  woidd  have  to  have  an 
unusually  high  number  of  seedlings  to  end  up  with  a  rea- 
sonably stocked  stand.  If  10  to  20  percent  are  resistant, 
the  probability  of  producing  a  new  stand  would  be  consid- 
erably better. 


Figure  5 — Proliferated  lateral  buds  on  whitebark  pine. 


123 


Locate  Parents 
(Resistant  Phenotypes) 

Collect  Wind  or 
Cross-Pollinated  Seed 

Blister  Rust  Test 
1 0  Years 


Graft  Seed  Orchard 
Production 
1 0  Years 


Seedling  Seed  Orchard 
Production 
40  Years 


Outplant  Survivors 
From  Blister  Rust 
Test 


Stand  Establishment 
50  Years 


Stand  Establishment 
50  Years 


Stand  Establishment 
50  Years 


Total  Time 
70  Years 


Total  Time 
100  Years 


Total  Time 
60  Years 


Figure  6 — Options  for  traditional  breeding  methods  for  developing  new  varieties  of  blister  rust-resistant 
whitebark  pine,  including  completion  times. 


2.  How  much  inbreeding  is  occurring  in  the  decimated 
stands?  Severe  inbreeding  might  result  in  reduced 
growth  and  survivabihty. 

3.  What  is  the  minimimi  number  of  cones  £ind  seed 
needed  before  seed  caches  are  effective  for  regeneration? 
Some  sites  may  be  beyond  help. 

4.  When  these  sites  start  producing  resistant  seed  how 
far  can  the  seed  be  transferred? 

Timing  (fig.  8):  First-generation  blister  rust  selection 
has  already  been  made  on  parent  trees.  Seedling  estab- 
lishment has  already  occurred  on  dozens  of  sites.  Natursd 
selection  by  blister  rust  is  occmring.  In  40  to  50  years  es- 
tablished stands  on  preferred  sites  could  be  producing 
cones. 

However,  it  is  not  only  the  increase  of  resistance  genes 
that  is  important — although  this  is  a  key  to  the  species' 
future — it  is  important  to  maintain  as  much  of  the  natu- 
ral function  of  whitebark  pine  forests  as  possible.  In  addi- 
tion to  the  severe  genetic  impact  to  whitebark  pine  itself, 
the  critical  ecological  functions  of  this  species  have  been 
threatened  by  loss  of  integrity  of  whitebark  pine  forests. 
Thresholds  are  a  key  consideration  when  attempting  to 
restore  severely  altered  ecological  systems  (Friedel  1991; 
Laycock  1991).  The  ecological  thresholds  for  whitebark 
pine  forests  are  not  well  understood  and  with  the  loss  of 
most  of  the  original  structure  of  these  forests  much  of  the 
integrity  has  been  lost.  It  will  take  generations  to  repro- 
duce nearly  pure,  mature  whitebark  pine  forests.  In  the 
interim,  components  of  the  ecological  systems  most  af- 
fected by  the  loss  of  whitebark  pine  forests  may  be  irre- 
versibly altered. 

Time  is  of  the  essence,  but  all  is  not  lost.  Over  large  ar- 
eas the  first  generation  is  mostly  gone,  but  in  many  areas 


the  second  generation  is  there;  we  must  act  now  to  save  it. 
In  other  sireas  the  first  generation  is  largely  intact.  In 
these  areas  immediate  emphasis  needs  to  be  placed  on  re- 
taining or  restoring  the  ecological  function  of  whitebark 
pine.  Protective  measures,  such  as  pruning  and  Ribes 
population  manipidation  may  be  used. 

What  we  are  trying  to  say  in  this  section  is  that  there  is 
much  that  can  and  must  be  done  immediately  to  maintain 
and  restore  whitebark  pine  in  the  ecosystem. 

BLISTER  RUST  ON  WHITEBARK 
PINE,  A  GENETICS  PERSPECTIVE 

The  general  level  of  resistance  of  the  original  popula- 
tion (before  the  introduction  of  blister  rust)  is  less  than 
1  percent.  Therefore,  the  genetic  consequence  of  the  blis- 
ter rust  epidemic  for  most  of  the  range  of  whitebark  pine 
is  to  push  it  through  an  extremely  small  bottleneck.  In 
some  areas  where  the  number  of  surviving  trees  has  been 
reduced  to  only  a  few  over  several  hundred  hectares,  the 
bottleneck  may  result  in  entire  populations  being  lost. 
Figure  1  indicates  geographic  populations  of  whitebark 
pine  that  will  have  the  most  immediate  difficulty  (high  in- 
cidence area),  those  that  will  have  only  minor  problems 
(moderate  incidence  area),  and  those  that  are  not  likely  to 
have  a  problem  (little  or  no  incidence).  Fiulher  research 
clarifying  levels  of  natural  resistance  will  be  a  significant 
aid  to  predicting  the  adequacy  of  natural  selection  in  the 
various  areas.  On  some  sites,  time  alone  may  return 
whitebark  pine  to  its  former  levels;  on  other  sites,  white- 
bark pine  may  be  lost  if  not  aided  by  genetic  improvement 
and  silvicultural  programs. 

In  areas  where  there  are  many  surviving  trees  within  a 
very  reduced  popidation,  the  concern  will  be  the  damage 


124 


Whitebark  Pine  Competing  Species 

Figure  7 — Development  of  a  natural  selection  stand:  A,  A  clearcut  before  cleaning  containing  whitebark 
pine  and  competing  trees;  B,  clearcut  after  cleaning;  C,  clearcut  after  infection  by  blister  rust  and  selection 
of  most  resistant  trees;  D,  the  natural  selection  stand  or  seed  orchard  after  growth  of  selected  trees. 


Locate  Sites 
Remove  Competing 
Trees 


Stand  Establishment 
50  Years 


Some  stands  are 
already  20  years  old 


Total  Time 
50  Years  or  Less 


Figure  8 — Natural  selection  method  for  developing 
a  new  vahety  of  blister  rust-resistant  whitebark  pine, 
including  completion  times. 


done  by  inbreeding.  In  most  conifers  inbreeding  is  detri- 
men1:al.  Inbred  trees  generally  grow  slowly,  and  detrimen- 
tal genes  such  as  for  albinism,  short  leaves,  and  stunted 
seedlings  may  often  occur.  However,  whitebark  pine  is  of- 
ten transported  by  birds  to  areas  where  there  is  no  other 
whitebark  pine  (Linhart  and  Tomback  1985).  Because 
there  are  only  a  few  individuals  in  the  start-up  stand,  in- 
breeding would  be  high.  This  may  have  been  the  most 
common  way  of  past  population  expansion  and,  over  time, 
lethal  and  other  detrimental  genes  may  have  dropped  out 
of  the  population  leaving  a  species  that  is  not  adversely 
affected  by  inbreeding. 

Even  vTith  all  these  dire  genetic  consequences,  the  ge- 
netic opportimity  for  the  survival  of  whitebark  pine  in  the 
high-mortality  zone  appears  very  high,  especially  with  the 
intelHgent  help  of  humans.  There  are  many  phenotypi- 
cally  resistant  trees.  These  surely  can  be  used  as  a  base 


125 


for  new  varieties  of  whitebark  pine  resistant  to  blister 
rust.  But  most  exciting,  the  high  natural  regeneration  of 
whitebark  pine  stands  in  clearcuts  and  burns — even  in  ar- 
eas where  there  appear  to  be  insufficient  nimibers  of  par- 
ent trees — indicates  that  natural  processes  are  already 
producing  a  new  variety.  Further,  these  trees  are  grow- 
ing well;  there  do  not  appear  to  be  inbreeding  problems. 

CONCLUSIONS 

The  consequences  of  the  blister  rust  epidemic  have  to  be 
viewed  as  a  disaster  for  whitebark  pine,  and  a  setback  for 
Clark's  nutcrackers,  grizzly  bears,  and  other  components 
of  whitebark  pine  forest  ecosystems.  Nevertheless,  white- 
bark pine  appears  to  be  equipped  with  elements  that  will 
permit  survival.  Most  important  is  seed  caching  by  the 
nutcracker  even  when  there  is  just  a  small  excess.  Also, 
phenotypically  resistant  trees  seem  to  be  fairly  common. 
Considering  the  low  numbers  of  noncankered  western 
white  pine  trees  together  with  the  much  higher  suscepti- 
bility of  whitebark  pine,  we  wonder  that  there  are  any 
whitebark  pine  trees  at  all. 

Over  large  areas,  the  first  generation  of  whitebark  pine 
is  nearly  gone;  however,  the  second  generation  is  often 
present.  In  other  areas  the  first  generation  is  still  largely 
intact.  To  save  whitebark  pine  and  its  ecosystem  in  the 
most  sensitive  areas  it  is  essential  that  we  act  now.  Main- 
taining sufficient  ecosystem  function  for  whitebark  pine 
while  developing  greater  resistance  will  be  a  challenge  for 
land  managers. 

REFERENCES 

Arno,  S.  F.;  Weaver,  T.  1990.  Whitebark  pine  community 
types  and  their  patterns  on  the  landscape.  In:  Schmidt, 
W.  C;  McDonald,  K  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-moimtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  97-105. 

Bedwell,  J.  L.;  Childs,  T.  W.  1943.  Susceptibility  of  white- 
bark pine  to  blister  rust  in  the  Pacific  Northwest.  Jour- 
nal of  Forestry.  41:  904-912. 

Bingham,  R.  T.  1972a.  Artificial  inoculation  of  large  num- 
ber of  Pinus  monticola  seedlings  with  Cronartium 
ribicola.  In:  Bingham,  R.  T.;  Hofi",  R.  J.;  McDonald, 
G.  I.,  eds.  Biology  of  rust  resistance  in  forest  trees. 
Misc.  Publ.  1221.  Washington,  DC:  U.S.  Department  of 
Agriculture:  357-372. 

Bingham,  R.  T.  1972b.  Taxonomy,  crossability,  and  rela- 
tive blister  rust  resistance  of  5-needled  white  pines.  In: 
Bingham,  R.  T.;  Hoff,  R.  J.;  McDonald,  G.  I.,  eds.  Biol- 
ogy of  rust  resistance  in  forest  trees.  Misc.  Publ.  1221. 
Washington,  DC:  U.S.  Department  of  Agric;ilture: 
271-280. 

Bingham,  R.  T.  1983.  Blister  rust  resistant  western  white 
pine  for  the  Inland  Empire:  the  story  of  the  first  25 
years  of  the  research  and  development  program.  Gen. 
Tech.  Rep.  INT-146.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermovmtain  Forest  and 
Range  Experiment  Station.  45  p. 


Childs,  T.  W.;  Bedwell,  J.  L.  1948.  Susceptibility  of  some 
white  pine  species  to  Cronartium  ribicola  in  the  Pacific 
Northwest.  Journal  of  Forestry.  46:  595-599. 

Childs,  T.  W.;  Bedwell,  J.  L.;  Englerth,  R.  H.  1938.  Bhster 
rust  infection  of  Pinus  albicaulis  in  the  Northwest. 
Plant  Disease  Reporter.  22:  139-140. 

Delatour,  C;  Birot,  Y.  1982.  The  international  lUFRO  ex- 
periment on  resistance  of  white  pines  to  blister  rust 
{Cronartium  ribicola):  the  French  trial.  In:  Heybroek, 
H.  M.;  Stephan,  B.  R.;  Weissenberg,  K.,  eds.  Resistance 
to  diseases  and  pests  in  forest  trees.  Wageningen: 
Pudoc:  412-414. 

Friedel,  M.  H.  1991.  Range  condition  and  the  concept  of 
thresholds:  a  viewpoint.  Journal  of  Range  Management. 
44:  422-426. 

Hoff,  R.  J.;  Bingham,  R.  T.;  McDonald,  G.  I.  1980.  Relative 
blister  rust  resistance  of  white  pines.  European  Journal 
of  Forest  Pathology.  10:  307-316. 

Laycock,  W.  A.  1991.  Stable  states  and  thresholds  of 
range  condition  on  North  American  rangelands:  a  view- 
point. Journal  of  Range  Management.  44:  427-433. 

Linhart,  Y.  B.;  Tomback,  D.  F.  1985.  Seed  dispersal  by 
nutcrackers  causes  multi-trunk  growth  form  in  pines. 
Oecologia.  67:  107-110. 

McCaughey,  W.  W.;  Schmidt,  W.  C.  1990.  Autecology  of 
whitebark  pine  (Pinus  albicaulis  Engelm.).  In:  Schmidt, 
W.  C;  McDonald,  K  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  85-96. 

McDonald,  G.  I.;  Hoff,  R.  J.  1970.  Resistance  to  Cro- 
nartium ribicola  in  Pinus  monticola:  early  shedding  of 
infected  needles.  Res.  Note  INT-124.  Ogden,  UT:  U.S. 
Department  of  Agriculture,  Forest  Service,  Intermoun- 
tain  Forest  and  Range  Experiment  Station.  8  p. 

Pfister,  R.  D.;  Kovalchick,  R.  L.;  Arno,  S.;  Presby,  R.  1977. 
Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Forest  and  Range 
Experiment  Station.  174  p. 

Spaulding,  P.  1911.  The  blister  rust  of  white  pine.  Bull. 
206.  Washington,  DC:  U.S.  Department  of  Agriculture, 
Bureau  of  Plant  Industry.  88  p. 

Steele,  R.;  Cooper,  S.  V.;  Ondov,  D.  M.;  Roberts,  D.  W.; 
Pfister,  R.  D.  1983.  Forest  habitat  t3rpes  of  eastern 
Idaho-western  Wyoming.  Gen.  Tech.  Rep.  INT-144. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Forest  and  Range  Experiment  Sta- 
tion. 122  p. 

Steele,  R.;  Pfister,  R.  D.;  Ryker,  R.  A.;  Kittams,  J.  A.  1981. 
Forest  habitat  types  of  central  Idaho.  Gen.  Tech.  Rep. 
INT-114.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  138  p. 

Stephan,  B.  R.  1985.  Resistance  of  five-needle  pines  to 
blister  rust.  Allgemeine  Forstzeitschrift.  No.  28: 
695-697. 

Yokota,  S.  1983.  Resistance  of  improved  Pinas  monticola 
and  some  other  white  pines  to  the  blister  rust  fungus, 
Cronartium  ribicola,  of  Hokkaido,  Japan.  European 
Journal  of  Forest  Pathology.  13:  389-402. 


126 


PERFORMANCE  OF  PINUS  CEMBRA, 
P.  PEUCE,  AND  P.  STROBIFORMIS 
WITHIN  AIR-POLLUTED  AREAS 

Karel  Kahak 


Abstract — ^An  inventory  of  the  surviving  species  within  the 
heavily  air-polluted  area  of  the  Ore  Mountains  showed  a  number 
of  species  of  the  genus  Pinus  without  symptoms  of  injury.  Three 
species  of  the  stone  pines  are  described  to  demonstrate  the  im- 
portance of  the  geological  past  of  the  species.  It  relates  to  the 
impact  of  the  catastrophes  and  volcanic  activities  and  sudden 
changes  of  climate  for  the  species'  preadaptation  to  the  toxic 
impact  of  a  changing  environment.  The  causes  of  the  resistance 
seemed  to  be  connected  with  the  long-term  migration  {Pinus 
cembra,  P.  peace)  and  hybrid  origin  (P.  peace,  P.  strobiformis). 
The  importance  of  the  life  history  of  the  species  is  unquestioned. 


During  the  years  1956  until  1966  the  author  (Kahak 
1971,  1988,  1991)  estabhshed  an  international  collection 
of  species  of  gentis  Pinus  in  Plzeh,  West  Czechslovakia, 
in  the  forest  locality  of  Sofronka  (latitude  50°  N.). 

Various  species  of  pines  are  represented  in  Sofronka  ei- 
ther by  patterns  of  populations  that  come  from  different 
localities  within  every  species  range  or  from  groves  of  in- 
dividual trees  from  species  that  originated  from  locations 
that  were  not  sufficiently  preadapted  for  the  local  condi- 
tions of  the  new  environment  (Kahak  1988).  A  research 
center  was  established  at  the  Arboretxim  Sofronka  in  1963 
to  investigate  evolution  of  the  species  included  within  the 
collection.  The  entire  range  of  provenance  experiments 
with  some  of  the  species  was  analyzed  with  respect  to  the 
evolution  of  their  distribution  area  during  postglacial 
time. 

The  air  pollution  caused  by  extensive  industrial  emis- 
sions resulted  in  a  mass  extinction  of  the  Norway  spruce 
forests  in  the  Ore  Mountains  in  the  vicinity  of  Plzeh.  In 
the  first  stage  of  the  extinction,  the  dying  spruce  stands 
were  cleared,  but  small  groves  of  individual  trees,  pre- 
dominantly the  local  and  exotic  species  of  pines,  survived. 
Therefore,  the  staff  of  the  research  center  at  Sofronka  was 
invited  to  join  the  search  for  an  ecologically  harmless 
method  of  reconstructing  the  extinct  forests. 

Some  species  of  white  pines  were  discovered  whose 
needles  appeared  to  be  unaffected  by  emissions.  These 
species  of  stone  pines  were:  Swiss  stone  pine  {Pinus 
cembra  L.),  Macedonian  pine  (Pinus  peuce  Griseb.),  and 
southwestern  white  pine  {Pinus  strobiformis  Engelm.). 
Their  unique  viability  had  helped  them  to  survive  under 
the  highly  unfavorable  conditions. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Karel  Ka&fik,  National  Park  of  Simiava,  Vimperk,  Czechoslovakia. 


SWISS  STONE  PINE 

1.  Occurs  as  individual  trees  in  the  entfre  emission 
area  between  the  town  of  Hora  Sv.  Sebestiana  and  Mount 
Klmovec  (1,210  m  above  sea  level  [a.s.l.]).  Some  trees, 
otherwise  injured  by  snow  and  icing,  show  no  injury  from 
emissions.  Many  of  them  bear  germinant  seeds  (Mottl 
and  Prudic  1982). 

2.  One  grove  of  fertile  trees  60  to  70  years  of  age  occurs 
in  the  cemetery  arboretiun  in  the  town  of  Vejprty.  Their 
dimensions  are  12  cm  diameter  at  1.3  m  and  9-11  m 
height.  They  bear  germinant  seeds. 

MACEDONIAN  PINE 

On  the  Saxon  side  of  the  mountains: 

1.  Forest  district  Steinbach,  Department  157a2,  eleva- 
tion 800  m  a.s.l.  A  group  of  17  trees,  age  of  55  years, 
fertile,  with  natural  regeneration.  Dimensions:  Diameter 
26-35  cm  at  1.3  m  and  15  m  height. 

2.  Forest  district  Schmalzgrube,  Department  291a3, 
elevation  900  m  a.s.l.  Twelve  trees  of  the  same  prove- 
nance and  age  as  the  latter  growing  on  a  peat  bog!  The 
trees  are  healthy  and  fertile.  Dimensions:  Diameter 
15-40  cm  at  1.3  m  and  11-14  m  height.  Tree  tnmks  have 
been  injured  by  the  red  deer  and  humans  (chmbing 
sptars). 

3.  Forest  district  Grumbach,  Department  331a4,  ele- 
vation 825  m  a.s.l.  Stand  area  is  0.75  ha  and  trees  are 
50  years  old  (1982)  and  grow  in  a  mixture  with  Pinus 
contorta  ssp.  latifolia  and  the  movmtain  variant  of  Pinus 
sylvestris  L.  Average  height  of  the  stand  is  18  m,  diam- 
eter is  21-35  cm  at  1.3  m.  Macedonian  pine  is  the  most 
vital  of  the  named  species. 

On  the  Czech  side  of  the  mountains: 

1.  Locality  Bludna,  Forest  Management  Horm'  Blatna, 
elevation  1,000  m  a.s.l.  Northern  slope  stand  of  approxi- 
mately 200  trees,  age  probably  70  years.  Origin  tmknown, 
but  probably  from  activities  of  the  Forest  Research  Insti- 
tute Mariabnmn  just  before  World  War  I.  Many  trees  were 
apparently  broken  at  the  height  of  5-7  m,  but  replacement 
crowns  are  now  reaching  heights  of  15  m.  Trees  are  rela- 
tively fertile  with  natural  regeneration  present. 

SOUTHWESTERN  WHITE  PINE 

1.  Locality:  Cemetery  arboretum  in  the  town  of  Vejprty 
with  a  heavy  emission  impact.  One  tree  only.  Diameter 
36  cm  at  1.3  m,  height  14  m,  and  age  50-60  years.  Tree  is 
fertile  with  germinant  seeds  in  the  cones  Ijdng  on  the 
moist  grass! 


127 


PROBABLE  CAUSES  OF  RESISTANCE 
Autochtonous  Species 

Ten  years  of  observations  and  experimental  activi- 
ties in  the  7-year  period  1982-89  give  us  the  following 
indications: 

Variability — ^Among  the  local  forest  tree  species,  there 
are  different  degrees  of  resistance  between  individual 
trees  within  the  stands.  Both  autochtonous  and  alloch- 
tonous  pine  species  had  been  gradually  reduced,  but  there 
are  some  examples  without  any  symptoms  of  injury.  Fir 
disappeared  on  the  Czech  side  of  the  mountain  range 
first,  but  drought  rather  than  emissions  may  have  been 
the  cause.  Norway  spruce  seems  to  be  extinct. 

Catastrophic  Selection — Norway  spruce  mostly  dis- 
appeared in  a  relatively  short  time — about  6  years.  How- 
ever, after  its  massive  extinction,  individually  resistant 
trees  of  this  species  were  found  within  the  extinct  forests. 
This  appearance  is  commonly  known  under  the  term 
"population  bottleneck"  (Dobzhanski  and  others  1976)  or 
"catastrophic  selection"  (Lewis  1962;  Raven  1964)  or  "une 
evolution  acceleree  (Bouvarel  1960). 

Founder  Effect — ^According  to  Mayr's  "founder-effect" 
(Mayr  1942,  1965,  1979)  after  the  strong  reduction  of 
Norway  spruce  the  reconstruction  of  its  genetic  architec- 
ture within  the  small  number  of  survivors  (founder  popu- 
lation) might  follow  (Kahak  1988).  This  might  give  rise 
to  a  new  species  population  endemic  to  the  environmental 
conditions  that  caused  the  latter  extinction. 

Allochtonous  Species 

Hybridization — The  Canadian  species  of  spruce  show 
a  high  degree  of  emission  resistance  compared  to  the  local 
Norway  spruce.  Their  life  history,  with  their  intricate 
genetic  relations,  may  help  explain  this  orientation 
(Heimburger  and  others  1983;  Morgenstern  and  Farrar 
1964).  Morphological  and  genetic  evidence  indicates  that 
black  spruce  is  a  relatively  young  species.  It  is  thought  to 
be  of  an  American  origin  resulting  from  the  hybridization 
of  a  proto- white  spruce  from  eastern  Asia  migrating  east- 
ward with  a  proto-red  spruce  of  North  American  origin 
migrating  westward.  Differences  in  morphological  char- 
acteristics between  black  spruce  and  red  spruce  in  the  di- 
rection of  white  spruce  would  indicate  such  hybrid  origin 
(Heimburger  1983). 

Directional  Selection 

The  most  relevant  information  came  from  all  the  for- 
eign species  formerly  introduced  to  this  mountain  range. 
An  excellent  example  is  the  life  history  of  lodgepole  pine 
as  described  by  Hansen  (1942,  1943).  This  species  came 
from  the  Cascade  Range  territories  of  North  America 
where  it  had  been  influenced  by  repeated  volcanic  activ- 
ity. Consequently,  many  generations  of  lodgepole  pine 
were  subjected  to  directional  selection  through  the  toxic 
air  and  soil,  and  lodgepole  pine  adapted  to  become  the 
founding  pioneer  stage  of  the  ecosystems'  evolution. 


Migration  Preadaptation 

The  other  very  important  evolutionary  factor  is  migra- 
tion. It  acts  on  the  development  of  ecological  potentisds 
as  a  consequence  of  the  tolerance  to  the  extreme  envi- 
ronmental conditions  (toxic  sites),  the  environmental 
changes,  and  all  the  preadaptations  that  could  be  ac- 
quired only  during  a  long  migration  route  such  as  those 
in  the  Tertiary — one  from  eastern  Asia  to  southern 
Europe  (Mirov  1967)  and  the  other  from  eastern  Asia 
to  Australia  (Mayr  1965). 

HISTORY  OF  SPECIES 

Mirov  (1967)  located  the  center  where  the  genus  Pinus 
originated  in  Eastern  Asia,  where  catastrophes  like  vol- 
canic activity  and  changes  in  climate  gave  all  the  species 
of  pines  a  specific  starting  point  to  their  principal 
adaptation. 

"Pines  however  possessed  two  characteristics  since  the 
Mezozoic  origin  of  the  genus:  their  xeromorphy  and  their 
ability  to  endure  direct  sunlight.  These  characteristics 
may  be  considered  as  generally  the  same  throughout  the 
whole  paleobotanical  history."  (Mirov  1967). 

Pinus  cembra 

"During  the  Ice  Age,  the  high  mountain  pines  of  the 
Alps  either  perished  or  were  preserved  in  the  lower  areas 
not  covered  with  ice.  Pinus  cembra  found  occasionally  in 
northern  Italy  is  the  relic  preserved  in  one  of  such  refugia 
(Emberger  1944).  With  subsequent  warming  up,  it  mi- 
grated again  to  the  high  elevations  of  the  central  Alps. 
There  is,  of  course,  a  possibility  of  the  existence  of  non- 
glaciated  refugia  even  in  the  central  parts  of  the  Alps. 
Pinus  cembra  might  have  survived  in  some  of  these  few 
areas  that  remained  free  from  the  ice."  (Mirov  1967). 

The  indications  are  that  P.  cembra  and  many  other  spe- 
cies of  the  eastern  Asia  flora  migrated  to  the  Alps  during 
the  Tertiary.  During  the  last  glaciation  they  occurred  in 
an  almost  continuous  belt  of  higher  elevations.  Conse- 
quently, because  of  the  enlarged  area  of  its  distribution, 
P.  cembra  developed  a  well-supported  degree  of  genetic 
variability.  This  species  of  stone  pine  survived  the  last 
glaciation,  and  within  its  refugias  there  was  probably  an 
increasing  frequency  of  autogamy  that  must  be  taken  into 
account  (Bannister  1965).  After  the  retreat  of  the  glacier, 
the  migrating  progenies  probably  intercrossed  at  their 
contact  zone  and  thus  acquired  their  increasing  variabil- 
ity as  well.  Together  with  the  very  short  period  of  the 
last  glaciation,  the  imdiminished  variability  still  main- 
tains itself  today.  Pinus  peuce  apparently  has  a  similar 
pattern  of  the  life  history. 

Pinus  peuce 

According  to  Mirov,  "There  are  indications  that  Medi- 
terranean pines  migrated  from  eastern  Asia  along  the 
mountain  ranges,  that  once  extended  north  of  and  parallel 
to  Himalaya."  (Mirov  1967). 

"Endemic  Pinus  peuce  is  very  local  in  Bulgaria,  Albania 
and  Yugoslavia.  It  is  a  haploxylon  pine  closely  related  to 


128 


Pinus  griffithii  of  the  Himalaya.  Before  the  great  migra- 
tion, Pinus  pence  has  been  crossed  with  other  haploxylon 
pines.  Judging  by  the  chemical  composition  of  its  tiirpen- 
tine  (Illoff  and  Mirov  1956)  it  has  a  close  affinity  to  east- 
em  Asiatic  and  western  American  pines"  (Mirov  1967). 

The  hybridization  should  be  a  source  of  variation  for  ad- 
aptation to  new  environments  (Lewontin  and  Birch  1966). 

The  whole  Mediterranean  region  experienced  many  tec- 
tonic changes  throughout  the  Tertiary.  In  the  Quarter- 
nary  there  occurred  intensive  faulting  and  sinking  of 
considerable  areas.  This  resiilted  in  the  formation  of  sep- 
arate moimtain  ranges  and  many  islands.  As  a  conse- 
quence, the  present  distribution  of  pines  in  the  Mediter- 
ranean region  is  sporadic  and  irregular  (Mirov  1967). 

Pinus  strohiformis 

Southwestern  white  pine  appear  to  have  morphological 
indications  of  a  hybrid  origin.  Critchfield  and  Little 
(1966)  said:  "This  species  forms  a  link  both  geographi- 
cally and  morphologically,  between  the  neighbors  Pinus 
flexilis  and  Pinus  ayacahuite.  Southwestern  pine  is  ap- 
parently separated  from  its  southern  relative  Pinus  aya- 
cahuite by  a  100  mile  gap  that  does  not  seem  to  corre- 
spond to  any  comparable  discontinuity  in  suitable 
habitats." 

We  have  popvilation  samples  of  Pinus  strohiformis  in 
our  arboretum  that  originated  from  Greenlee,  AZ,  at  lati- 
tude 33°39'  N.,  longitude  109°16'  W.,  elevation  2,740  m 
a.s.l.  This  stand  is  now  30  years  of  age.  Its  fertility  began 
at  the  age  of  22  years.  Its  cones  are  surprisingly  variable 
in  size,  from  the  smaller  cones  similar  to  Pinus  flexilis  to 
very  big  cones  with  apophyses,  similar  to  Pinus  aya- 
cahuite var.  brachyptera.  Such  a  variation  supports  the 
hypothesis  about  the  hybrid  origin  of  this  species. 

TESTS  OF  SPECIES  POTENTIALS 

The  response  of  a  species  to  the  impact  of  an  environ- 
mental stress  allows  us  to  recognize  intrinsic  characteris- 
tics and  appearances  that  we  do  not  usually  encounter  un- 
der normal  circumstances.  Therefore,  we  are  focused  on 
testing  the  collection  of  pine  species  at  the  arboretum 
within  the  zone  where  there  is  a  heavy  impact  of  both 
emissions  and  severe  mountain  climates. 

CONCLUSIONS 

The  threatened  trees  of  some  forest  species  imder  heavy 
emissions  often  show  a  surprising  ability  to  endure  the  se- 
lection pressure  of  the  toxic  environment.  Evidently  this 
depends  on  the  life  history  of  the  species  in  question. 
Some  conditions  under  which  we  feel  the  tested  species 
have  a  better  predisposition  to  siirvive  are: 

1.  The  species  is  of  hybrid  origin. 

2.  The  species  has  imdergone  a  long-term  interconti- 
nental migration. 

3.  The  species  has  been  subjected  to  repeated  strong 
directional  or  even  catastrophic  selection  under  the  im- 
pact of  volcanic  activity.  Volcanic  areas  are  noted  for 
their  toxic  air  and  soil  similar  in  some  ways  to  industrial 
emission  areas. 


REFERENCES 

Bannister,  M.  H.  1965.  Variation  in  the  breeding  system 
of  Pinus  radiata.  In:  Baker,  H.  G.;  Stebbins,  G.  L.,  eds. 
The  genetics  of  colonizing  species.  In:  Proc.  First  Int. 
Un.  Biol.  Sci.  Symp.  Gen.  Biol.  Davis;  1964  February. 
New  York  and  London:  Academic  Press:  353-374. 

Bigelow,  R.  S.  1965.  Hybrid  zones  and  reproductive  isola- 
tion. Evolution.  19:  449-508. 

Bouvarel,  P.  1960.  Note  sur  la  resistance  au  froid  de 
quelques  provenances  de  pin  maitime.  Revue  forestiere 
francaise.  7:  495-508. 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture, 
Forest  Service.  97  p. 

Diels.  1910.  ex  Mirov  (1967). 

Dobzhanski,  Th.;  Ayala,  F.;  Stebbins,  L.  G.;  Valentine, 

L.  W.  1976.  Evolution.  San  Francisco:  Freeman. 
Emberger.  1944.  ex  Mirov  (1967). 

Hansen,  H.  1942.  The  influence  of  volcanic  eruptions  upon 

post-Pleistocene  forest  succession  in  central  Oregon. 

American  Journal  of  Botany.  29:  214-219. 
Hansen,  H.  1943.  Paleoecology  of  the  sand  dune  bogs  on 

the  southern  Oregon  coast.  American  Joiirnal  of  Botany. 

30:  335-340. 

Heimburger,  C.  C.  1983.  The  evolution  of  Black  spruce. 

Proc.  XDC.  Meet.  CTIA,  Part  2:  163-166.  Toronto. 
Illoff,  P.  M.,  Jr.;  Mirov,  N.  T.  1956.  Composition  of  gimi 

turpentines  of  pines.  XXV.  A  report  on  two  white  pines: 

Pinus  koraiensis  from  Korea  and  Pinus  peuce  from 

Macedonia.  Amer.  Pharm.  Ass.  J.  Sci.  Ed.  45:  77-81. 
Kahak  ,  K.  1971.  Arboretum  Sofronka.  Introduction 

results  in  1958-1968.  Praha.  38  p.  illus. 
Kahak,  K  1988.  Contribution  to  maintaining  continuity  of 

the  Norway  spruce  in  the  Ore  Moimtains.  Folia  Mus. 

Rer.  Natur.  Bohem.  Occid.  Plzai,  Botanika  27.  58  p. 
Kahak  ,  K  1991.  Colonizing  ability  of  some  introduced 

species  of  pines.  FoKa  dendrologica.  18:  187-200.  Veda, 

Bratislava. 

Lewis,  H.  1962.  Catastrophic  selection  as  a  factor  in  spe- 
ciation.  Evolution.  16:  257-271. 

Lewontin,  R.  C;  Birch,  L.  C.  1966.  Hybridization  as  a 
source  of  variation  for  adaptation  to  new  environments. 
Evolution.  20:  315-336. 

Mayr,  E.  1942.  Systematics  and  the  origin  of  species.  New 
York:  Columbia  University  Press. 

Mayr,  E.  1965.  Summary.  In:  Baker,  H.  G.;  Stebbins,  G. 
L.,  eds.  The  genetics  of  colonizing  species.  In:  Proc.  First 
Int.  Un.  Biol.  Sci.  Symp.  Gen.  Biol.  Davis;  1964  Febru- 
ary. New  York  and  London:  Academic  Press:  553-562. 

Mayr,  E.  1979.  Evolution  xmd  Vielfalt  des  Lebens. 
Springer-Verlag. 

Mirov,  N.  T.  1967.  The  Genus  Pinus.  New  York:  Ronald  Press. 

Morgenstem,  E.  K.;  Farrar,  J.  L.  1964.  Introgressive 
hybridization  in  red  spruce  and  black  spruce.  Tech. 
Rep.  4.  Faculty  of  Forestry,  Univ.  of  Toronto. 

Mottl,  J.,  Prudic,  Z.  1982.  Analyza  riistu  limby  na 
Klmovci — V  Krusnych  horach  a  jejf  roubavanci. 
Tschechisch.  Zpravy  les.  vyzkumu  27(4):  9-13. 

Parsons,  P.  A.  1963.  Migration  as  a  factor  in  natural  selec- 
tion. Genetica.  33:  184-206. 

Raven,  P.  1964.  Catastrophic  selection  and  edaphic 
endism.  Evolution.  17:  336-338. 


129 


COMPETITION  AND  CROWN 
CHARACTERISTICS  OF  WHITEBARK 
PINE  FOLLOWING  LOGGING  IN 
MONTANA,  U.S.A. 

Todd  Kipfer 
Katherine  Hansen 
Ward  McCaughey 


Abstract — This  study  was  designed  to  evaluate  crown  develop- 
ment of  whitebark  pine  (Pinus  albicaulis)  subjected  to  intertree 
competition  in  young  mountainous  stands  in  the  Yellowstone 
area.  A  distance-dependent  competition  index  was  used  to  assess 
intertree  competitive  effects  on  crown  growth.  Significant  in- 
verse correlations  were  found  between  the  competition  index  and 
height,  crown  diameter,  crown  diameter/height,  and  crown  vol- 
ume; however,  only  20  percent  of  the  variation  was  explained  by 
measured  variables.  Competition  thresholds,  indicating  when 
crown  growth  becomes  limited,  were  estimated.  In  future  stud- 
ies, crown  vigor  or  density  measures  could  perhaps  yield  a  better 
competition  indicator. 


This  study  addresses  the  response  of  regenerating 
whitebark  pine  {Pinus  albicaulis)  to  competition  following 
logging.  Human  activities,  such  as  logging,  in  the  movm- 
tainous  environments  of  western  North  America  are  cre- 
ating a  variety  of  impacts  at  both  the  individual  scale  and 
at  the  ecosystem  scale.  Concern  about  regeneration 
crown  growth  and  future  cone  production  in  logged  sites 
prompted  this  study.  Cone  production  is  a  product  of 
mean  shoot  production  multiplied  by  the  number  of  fertile 
shoots  (Weaver  and  Forcella  1986).  Effects  of  competition 
were  hypothesized  to  affect  crown  development  of  white- 
bark pine.  Intertree  competition  was  measured  on  indi- 
vidual, regenerating  whitebark  pine  in  logged  sites  in 
southwestern  Montana  near  Yellowstone  National  Park. 
This  study  is  part  of  a  larger  research  project  conducted 
by  Kipfer  (1992). 

Whitebark  pine  are  found  on  sites  with  a  wide  range 
of  geologic,  geomorphic,  and  ecologic  conditions  (Hansen- 
Bristow  and  others  1990).  Their  growth  characteristics 
depend  on  numerous  factors,  including  competition,  that 
vary  spatially  and  temporally.  Quantifying  the  influence 
of  competing  vegetation  on  conifer  grov^i^h  may  help  to  de- 
termine silvicultural  prescriptions  such  as  thinning  inten- 
sity, optimize  management  decisions,  and  provide  a  basis 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Todd  Kipfer  is  a  Doctoral  Student,  Department  of  Geography,  Arizona 
State  University,  Tempe,  AZ  85281;  Katherine  Hansen  is  an  Associate 
Professor  of  Geography,  Department  of  Earth  Sciences,  Montana  State 
University,  Bozeman,  MT  59717;  Ward  McCaughey  is  a  Research  For- 
ester, USDA  Forest  Service,  Intermountain  Research  Station,  Forestry 
Sciences  Laboratory,  Bozeman,  MT  59717. 


for  growth  models  of  young  forest  stands  (Wagner  and 
Radosevich  1991b).  For  example,  one  management  option 
for  whitebark  pine  is  to  increase  or  sustain  seed  produc- 
tion for  regeneration  and  for  a  food  source  for  the  Clark's 
nutcracker  (Nucifraga  columbiana),  the  red  squirrel 
{Tamiasciurus  hudsonicus),  the  grizzly  bear  (Ursus 
arctos  horribilis),  and  a  host  of  other  birds  and  mammals 
(Craighead  and  others  1982;  Kendall  1983;  Kendall  and 
Amo  1990;  McCaughey  and  Schmidt  1990).  Seed  produc- 
tion may  increase  with  v^dder,  more  diflFuse  crowns  (Eggers 
1986;  Spurr  and  Barnes  1980),  and,  therefore,  silvicultural 
prescriptions  that  increase  crown  diffiisivity  are  desirable. 
The  effects  of  silvicultural  management  options  for  main- 
taining or  enhancing  whitebark  pine  ecosystems  have  had 
limited  study  (Eggers  1990;  Schmidt  and  McCaughey 
1990). 

PREVIOUS  STUDIES 

Whitebark  pine  is  considered  relatively  intolerant  of  com- 
petition and  shade  (Amo  and  Weaver  1990),  although  the 
trees  may  be  more  shade  tolerant  at  an  early  stage  and 
less  tolerant  at  later  developmental  stages  (McCaughey 
and  Schmidt  1990).  Although  initial  observations  by 
Eggers  (1990)  indicate  that  suppressed  whitebark  pine 
seedlings  and  saplings  respond  little  to  the  removal  of  for- 
est competition,  the  effect  of  competition  on  postlogging 
regeneration  needs  further  study.  Weaver  and  others 
(1990)  assessed  competition  in  terms  of  stand  develop- 
ment and  hypothesized  that  although  the  density  of 
whitebark  pine  seedlings  remained  approximately  con- 
stant across  stands  of  different  ages,  growth  from  seedling 
to  sapling  size  could  not  be  supported  in  stands  older  than 
100  years.  They  suggested  that  clearings  needed  to  be 
larger  than  10  m^  for  seedlings  to  grow  to  sapling  size. 

The  effect  of  competition  on  the  growth  of  individual 
trees  has  been  studied  for  many  economically  important 
trees  in  plantations  or  homogeneous  stands  (Daniels  and 
others  1986;  Tome  and  Burkhart  1989;  Wagner  and 
Radosevich  1991a,b).  These  studies  have  primarily  used 
individual  tree  grovvi:h  models  based  on  competition  indi- 
ces and  have  shown  an  inverse  relationship  between  the 
amount  of  competition  from  neighbors  and  individual  tree 
growth  performance  (Daniels  1976;  Daniels  and  others 
1986;  Lorimer  1983;  Tome  and  Burkhart  1989).  This 
study  focused  on  crown  development  of  whitebark  pine 
as  related  to  a  distance-dependent  competition  index. 
Managers  could  use  this  index  to  determine  the  optimum 


130 


spacing  or  number  and  spatial  arrangement  of  competi- 
tors for  maximimi  crown  development  of  whitebark  pine, 

METHODS 

Three  logged  stands  were  sampled  in  the  Gallatin  Na- 
tional Forest  in  southwestern  Montana  near  Yellowstone 
National  Park.  The  stands  ranged  in  elevation  from  an 
average  of  2,260  m  at  Moose  Creek  to  2,475  m  and  2,535  m 
at  Teepee  Creek  1  and  2,  respectively.  The  three  stands 
had  been  clearcut  between  1968  and  1972.  Data  collec- 
tion focused  upon  intra-  and  intertree  characteristics  for 
whitebark  pine  and  neighboring  trees.  One  hundred 
whitebark  pine  individuals  (more  than  0.5  m  tall)  were 
identified  within  each  of  the  three  stands  by  a  systematic- 
random  sampling  procedure  along  transects.  Three 
different-sized  circular  plots  were  nested  around  each  in- 
dividual whitebark  pine.  Within  each  circular  plot,  trees 
of  specific  sizes  were  identified  as  competitors  using  the 
following  criteria:  within  the  3-m-radius  plot,  all  tree 
seedlings  and  saplings  were  identified  as  competitors; 
within  the  6-m-radius  plot  all  trees  with  a  diameter  at 
breast  height  (d.b.h.)  greater  than  or  equal  to  4  cm  and 
less  than  10  cm  were  identified  as  competitors;  and  within 
the  9-m-radius  plot  all  trees  with  d.b.h.  greater  than  or 
equal  to  10  cm  were  identified  as  competitors.  This 
method  approximated  a  fixed  angle  gauge  sweep  (such  as 
that  used  by  Tome  and  Burkhart  1989)  and  requires  that 
trees  farther  away  from  the  individual  tree  be  larger  to  be 
considered  competitors. 

Five  attributes  were  measured  for  each  whitebark  pine: 
(1)  age,  (2)  height,  (3)  diameter  at  breast  height  (d.b.h.), 
(4)  average  live  crown  diameter,  and  (5)  height  of  the 
maximum  crown  diameter  above  the  ground  surface. 
Crown  and  height  variables  were  measured  to  the  nearest 
0.1  m  and  d.b.h.  to  the  nearest  0.1  cm. 

Total  sample  size  was  reduced  from  300  to  220  after 
tree  ages  showed  that  some  selected  whitebark  pine  indi- 
cated some  had  established  prior  to  logging  (advance  re- 
generation). It  is  difficult  to  evaluate  intertree  competi- 
tion prior  to  logging,  and  these  were  therefore  eliminated 
from  further  study.  A  ratio  of  crown  diameter  to  tree 
height  iCD/H)  was  computed  to  reflect  the  degree  of  hori- 
zontal crown  growth.  Crown  volume  was  computed  using 
the  cone  volume  equation:  (Volume  =  1/3*  PI  *  R'^  *  H). 
Measurements  on  competing  trees  within  the  9-m-radius 
plot  were  tree  species,  total  tree  height,  diameter  at 
breast  height,  and  distance  (measured  to  nearest  0.1  m) 
to  the  subject  whitebark  pine. 

A  distance-weighted  size-ratio  index  or  distance- 
diameter  index  (Alemdag  1978;  Daniels  1976;  Hamilton 
1969;  Hegyi  1974;  Tome  and  Burkhart  1989)  was  selected 
to  develop  a  competition  index,  from  which  intertree  com- 
petition could  be  quantified.  The  distance-diameter  index 
is  a  distance-dependent  index  that  svuns  a  size  ratio  of 
diameters  (d.b.h.  or  basal  diameter)  of  competing  trees  to 
the  diameter  of  a  subject  tree.  The  general  theory  is  that 
larger  competing  trees,  relative  to  the  individual  tree, 
contribute  more  competition  than  do  smaller  trees.  The 
index  also  assumes  that  competitive  influence  decreases 
with  increasing  distance  between  the  individual  tree  and 
the  competitor. 


The  distance-diameter  index  was  modified  for  white- 
bark pine  because  the  trees  often  had  multiple  stems  or 
did  not  reach  breast  height.  Height  was  substituted  for 
diameter  and  distance  weighting  functions  were  applied 
to  develop  the  modified  competition  index: 

N 

C.I.=  Z  (if./i/.)*function(Z)/ST..) 
i=l    '  ' 

where  C.I.  is  the  competition  index;  if.  is  the  tree  height 
of  the  subject  tree;  H.  is  the  tree  height  of  the  competing 
tree;  and  DIST.^  is  a  distance  decay  function.  Three  dis- 
tance decay  functions  were  used:  {1/DIST  ,  l/DIST^  ,  and 
e-°'s^y). 

Regression  analysis  was  used  to  evaluate  the  relation- 
ship between  competition  and  crown  characteristics  of  the 
whitebark  pine.  The  three  distance-decay  functions  were 
used  as  dependent  variables  and  tree  height  and  three 
crown  measures  (crown  diameter,  crown  diameter  to  tree 
height  ratio,  and  crown  volimie)  as  independent  variables. 

RESULTS 

Competition  index  values  varied  according  to  the  form 
of  the  distance-decay  function  used  for  whitebark  pine. 
Index  values  calculated  using  1/distance  and  1/distance^ 
were  similar  with  mean  values  ranging  from  13.12  to 
17.12.  Index  values  using  the  e~^'^^'^^  decay  function 
ranged  from  3.11  to  4.22. 

Tree  heights  of  whitebark  pine  were  significantly  corre- 
lated (P  =  0.007)  with  the  competition  index;  however, 
values  for  the  three  distance-decay  functions  (0.033  for 
1/distance,  0.027  for  l/distance^,  and  0.032  for  e"^''*^^^) 
indicate  that  the  linearity  of  that  relationship  is  poor  as 
shown  for  1/distance  in  figure  1.  The  intercept  value  in 
the  regression  equation  (y  =  a  +  bx)  was  2.047,  which  may 
represent  maximimi  tree  heights  for  whitebark  pine  in 
the  age  classes  represented  by  this  study,  when  growing 
in  the  absence  of  competition. 

Crown  diameters  of  whitebark  pine  were  significantly 
correlated  {P  =  0.000)  with  the  competition  index,  shown 
for  1/distance  in  figure  1,  with  R^  values  varying  depend- 
ing on  the  form  of  the  distance-decay  function  (0.131  for 
1/distance,  0.096  for  l/distance^,  and  0.128  for  e"^*^^^^). 
The  intercept  value  (0.820)  of  the  regression  equation 
poorly  represents  the  maximum  possible  crown  diameter, 
since  crown  diameter  values  above  1.0  m  were  common. 

The  crown  diameter  to  tree  height  ratio  and  crown  vol- 
ume were  significantly  correlated  to  competition  index  as 
shown  for  the  1/distance  decay  function  in  figure  1.  Re- 
gression R^  values  for  crown  diameter  to  tree  height  ratio 
varied  depending  on  the  form  of  the  distance-decay  func- 
tion in  the  competition  index  equation  (0.122  for  1/distance, 
0.079  for  l/distance^,  and  0.120  for  e-^'^'=^).  Similar  regres- 
sion i?^  values  were  obtained  in  correlating  competition 
index  with  crown  volume  using  the  three  forms  of  the 
distance-decay  function  (0.092  for  1/distance,  0.058  for 
l/distance^  and  0.091  for  e-^^^-^^). 

The  1/distance  decay  function  in  the  competition  index 
model  yielded  slightly  higher  R  ^  values  for  tree  height, 
crown  diameter,  crown  diameter  to  tree  height  ratio,  and 
crown  voliraie  of  whitebark  pine.  The  1/distance  decay 


131 


S.0- 
4.5' 
4.0 

3.5  ■ 

ao 
^6 

2.0 

1J 

1.0 

as 
ao 


Tro  halght  »  2.047  -  0.0133'C.I. 
R-squarod  >  0.033 
p  =  0.007  n  >  220 


•  •       •  • 

£  li  *•*  t  .y  .  r 


* 


—  IX)- 


0.0 


Crown  diameter  >  0.820  -  0.0117*C.I. 
R-tquarad  =  0.131 
p  -  0.000  n  -  220 


•  a     •  a 

•  aa  « 

a  a 

•aaa  a  a  a  a 


aaa   •  a  •      a  a 
■       a  aa  a  a 
a        a  a 

a         a  a  aa 
•a  a    «a  a 


a  aa  •aaa 

•aa  a  aa  a« 

a     a        a  a 

aaa  a  aa  • 

a        aa>*  a 

a  a  aa  • a 

a  aaa 


a  a 
a  a 


a  a  a 
a  a 


Competition   Index  (I/distance) 


Competition  index  (l/dietance) 


a 

a 

Crown  diameter/Tree  height  «  0.428  -  0.00481 'CI. 
R-squarad  '  0.122 
p  -  0.000  n  3  220 

a 

Sa          •    •  a 
a'«       a-a  a 
J  a        a       •  •• 

a 

aaa       a  • 
a^    ^'V.-.a  ;C 

a  • 
•      a  •  •                      •  a 

a     a  a 

•  ^"'^ 

a 

aa 

• 

10  20  30  40  60 

Competition   index  (I/distance) 


60 


CO 

< 


Crown  vokima  <s  0.300  -  0.00056*0.1. 
R-aquarad  •  0.002 
p  «  0.000  n  =  220 


#  • 


•  t  • 


•••'a  *. 


10  20  30  40  SO 

Competition   index  (1/distance) 


Figure  1 — Bivariate  scatterplots  of  total  tree  heiglit,  crown  diameter,  crown  diameter/lielght,  and  crown 
volume  of  whiitebarl^  pine  plotted  against  competition  index  using  a  distance-decay  function  of  1/distance. 
Regression  values  shown  are  model  and  model  coefficients,  fl-square,  p-level,  and  sample  size  (n).  (Data 
from  all  tiiree  study  sites — iy/loose  Creek,  Teepee  Creek  1 ,  and  Teepee  Creek  2.) 


function  was  used  for  development  of  a  potential  competi- 
tion threshold  because  it  was  consistently  a  stronger  pre- 
dictor of  the  four  growth  variables. 

Scatterplots  were  evaluated  from  each  study  site  for 
competition  index  values  above  which  there  was  visually 
less  variability  in  the  data.  No  identifiable  threshold  was 
observed  for  whitebark  pine  in  the  Moose  Creek  stand, 
and  therefore,  in  order  to  increase  the  sample  size,  the 
two  Teepee  Creek  stands  were  combined.  Potential  com- 
petition thresholds  were  estimated  for  total  tree  height, 
crown  diameter,  crown  diameter  to  tree  height  ratio,  and 
crown  voltmie  for  sampled  whitebark  pine  in  the  two  Tee- 
pee Creek  stands  (fig.  2).  Variability  of  data  below  poten- 
tial competition  thresholds  indicates  that  tree  growth  fac- 
tors may  not  be  limiting.  The  majority  of  whitebark  had 
competition  index  values  below  the  estimated  threshold 
values. 

Regression  analysis  of  the  competition  index  against 
the  four  whitebark  pine  measures  using  only  those  trees 
greater  than  the  threshold  values  jdelded  poor  results  for 


tree  height  (R^  =  0.128,p  =  0.062,  n  =  27),  crown  diameter 
(R^  =  0.146,  p  =  0.106,  n  =  18),  and  crown  volume  (R^  = 
0.091,  p  =  0.120,  n  =  27).  A  statistically  significant  corre- 
lation, however,  was  found  between  the  competition  index 
and  the  crown  diameter  to  tree  height  ratio  (CD/H  = 
0.846  -  0.0142*C.I.:  R^  =  0.230, p  =  0.000,  n  =  18). 

DISCUSSION 

The  specific  competition  index  values  developed  by  this 
study  explain  approximately  20  percent  of  the  variation  in 
competitive  pressure  exerted  on  whitebark  pine  from  the 
numbers,  sizes,  and  spatial  patterns  of  neighboring  trees. 
Assessment  of  competition  on  crown  development  of 
whitebark  pine  is  difficult  when  data  come  from  present 
spatial  relationships  and  past  growing  conditions  are  un- 
known. Competition  was  a  significant  factor  influencing 
crown  characteristics  of  whitebark  pine  in  this  study.  It 
is  important  to  note  that  these  results  only  show  that 
competition  is  a  significant  factor  influencing  whitebark 


132 


5.0 
4.5 
4.0 

3.5 

ao 

2.5 
2.0 
1.5 
1.0 
OJ 

ao 


• 

• 

• 

• 

<               Potential  threshold 

• 

;i.r  J»  •  t,  - 

•  •  • 

•  •  • 

••  • 

*  *  • 

• 

•••      •      ••  • 

•    •  • 

•  • 

20  30  40  SO 

Competition    index  (1/distance) 


E 


OX 


(  •  •••• 


•  •  •  •  •  I 


Potential  threshold 


10  20  30  40  50 

Competition   Index  (1/dlstance) 


Potential  threshold 

• 

• 

<  

• 

\,     •  •  ' 

•  •  •  •       •  • 

•  • 

m  • 
• 

• 

• 

• 

•    •  • 
m 

•  • 

20  30  40  60 

Compstltlon  Index  (1/dlstance) 


n 
< 

o 

E 

o 
> 


•  •  •/ 


Potential  threshold 


10  20  30  40  £ 

Competition   index  (I/distance) 


Figure  2 — Potential  competition  thresholds  for  bivariate  scatterplots  of  total  tree  height,  crown  diameter, 
crown  diameter/height,  and  crown  volume  of  whitebark  pine  plotted  against  competition  index  using  a 
distance-decay  function  of  1/distance.  (Data  from  Teepee  Creek  1  and  Teepee  Creek  2  only.) 


pine.  The  mechanisms  of  this  competitive  influence  are 
unknown. 

The  distance-decay  function  1/distance  in  the  competi- 
tion index  model  created  an  expected  inverse  relationship 
of  tree  growth  measure  to  index.  As  distance  from  the 
subject  tree  increases  effects  from  competition  trees  de- 
crease. Tree  heights  and  crown  volumes  had  the  poorest 
correlations  with  the  competition  index.  Poor  correlations 
with  crown  volume  may  be  due  to  sampHng  procedures. 
In  future  studies,  crown  vigor  and  crown  biomass  meas- 
urements (stem  spacing)  might  be  used  to  weight  crown 
volume  measures,  perhaps  yielding  a  better  indication  of 
the  effect  of  competition  on  growth. 

Crown  diameters  and  crown  diameter  to  tree  height  ra- 
tios had  good  correlations  with  the  competition  index. 
Still,  only  12  to  13  percent  of  the  variation  in  these  meas- 
ures was  explained  by  the  competition  index.  The  addi- 
tion of  age  and  an  indicator  variable  for  stand  improved 
this  relationship  for  the  crown  diameter  measure,  increas- 
ing the  R  2  value  to  0.250.  It  did  not,  however,  yield  a  sig- 
nificant correlation  for  the  crown  diameter  to  tree  height 


ratio.  The  crown  diameter  to  tree  height  ratio  may  repre- 
sent a  characteristic  crown  shape  for  young  whitebark 
pine. 

Poor  correlations  between  whitebark  pine  crown  charac- 
teristics and  competition  index  may  be  due  to  factors  in- 
fluencing individual  tree  size  (local  density,  plant  geno- 
type, seed  size,  emergence  time,  microhabitat  variations, 
and  unknown  historical  growing  conditions).  A  competi- 
tion index  coxdd  incorporate  a  variety  of  growing  condi- 
tions over  time  and  the  use  of  incremental  growth  charac- 
teristics of  annual  growth  could  provide  some  improvement 
without  substantial  measurement  efforts. 

Potential  competition  thresholds  estimated  for  the  two 
Teepee  Creek  sites  may  be  artifacts  due  to  relatively  few 
trees  with  high  competition  values,  or  they  may  be  actual 
values  of  competitive  pressure  indicating  when  growth 
becomes  limited  by  competition.  Multiple-aged  stands 
with  varying  tree  densities  should  be  sampled  in  future 
studies.  To  compare  stands  of  different  ages,  the  age- 
independent  competition  index  methods  of  Lorimer  (1983) 
may  provide  a  strong  starting  point.  Although  regression 
analysis  can  comparatively  evaluate  the  influences  of 


133 


competitive  pressure  on  whitebark  pine  characteristics, 
actual  relationships  are  more  complex  than  the  linear  re- 
gression models  might  suggest.  Nonlinear  regression 
methods  should  be  explored  for  evaluating  competition 
influences. 

Timber  harvests  provide  an  important  but  declining 
part  of  our  regional  economy  (Powers  1991),  and  economic 
considerations  are  an  important  component  of  forest  man- 
agement policies.  Timber  harvest  can  cause  accelerated 
slope  failures,  erosion,  and  stream  sedimentation  from 
roads  (Marston  and  Anderson  1991)  when  improperly 
done.  Road  building  in  association  with  logging  provides 
access  to  whitebark  pine  regeneration  sites;  however,  it  is 
not  believed  that  logging  will  be  a  major  means  for  regen- 
erating whitebark  pine  (Arno  1986). 

It  is  imperative  to  find  viable  management  options  for 
increasing  crown  growth  and  the  potential  for  increased 
cone  production  for  whitebark  pine  because  of  its  impor- 
tance to  the  endangered  grizzly  and  many  other  wildlife 
species  in  the  Yellowstone  ecosystem.  The  importance  of 
quantifying  how  competing  trees  influence  crown  charac- 
teristics of  whitebark  pine  in  logged  stands  comes  from  a 
need  to  increase  our  knowledge  of  how  competition  affects 
crown  development  of  whitebark  pine. 

ACKNOWLEDGMENT 

This  research  was  supported  in  part  by  funds  provided 
by  the  Intermoxmtain  Research  Station,  Forest  Service, 
U.S.  Department  of  Agricultiu*e. 

REFERENCES 

Alemdag,  I.  S.  1978.  Evaluation  of  some  competition  indi- 
ces for  the  prediction  of  diameter  increment  in  planted 
white  spruce.  Inf.  Rep.  FMR-X-108.  Canadian  Forest 
Service  Forest  Management  Institute.  39  p. 

Arno,  S.  F.  1986.  Whitebark  pine  cone  crops — a  diminish- 
ing source  of  wildlife  food?  Western  Journal  of  Applied 
Forestry.  1(3):  92-97. 

Arno,  S.  F.;  Weaver,  T.  1990.  Whitebark  pine  community 
types  and  their  patterns  on  the  landscape.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposium  on  whitebark  pine  ecosystems:  ecology 
and  management  of  a  high-mountain  resource;  1989 
October  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  97-105. 

Craighead,  J.  J.;  Svunner,  J.  S.;  Scaggs,  G.  B.  1982.  A  de- 
finitive system  for  analysis  of  grizzly  bear  habitat  and 
other  wilderness  resources.  Wildlands  Inst.  Monogr.  1. 
Missoula,  MT:  University  of  Montana.  279  p. 

Daniels,  R.  F.  1976.  Simple  competition  indices  and  their 
correlation  with  annual  loblolly  pine  tree  growth.  Forest 
Science.  22:  454-456. 

Daniels,  R.  F.;  Burkhart,  H.  E.;  Clason,  T.  R.  1986.  A 
comparison  of  competition  measures  for  predicting 
growth  of  loblolly  pine  trees.  Canadian  Journal  of  For- 
est Research.  16:  1230-1237. 


Eggers,  D.  E.  1986.  Management  of  whitebark  pine  as  po- 
tential grizzly  bear  habitat.  In:  Contreras,  Glen  P.; 
Evans,  Keith  E.,  comps.  Proceedings — grizzly  bear  habi- 
tat symposiimi;  1985  April  30-May  2;  Missoula,  MT. 
Gen.  Tech.  Rep.  INT-207.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  170-175. 

Eggers,  D.  E.  1990.  Silvicultiiral  management  alterna- 
tives for  whitebark  pine.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  maneige- 
ment  of  a  high-mountain  resource;  1989  October  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  324-328. 

Hamilton,  G.  J.  1969.  The  dependence  of  volume  incre- 
ment of  individual  trees  on  dominance,  crown  dimen- 
sions, and  competition.  Forestry.  42:  133-144. 

Hansen-Bristow,  K.;  Montague,  C;  Schmid,  G.  1990. 
Geology,  geomorphology,  and  soils  within  whitebark 
pine  ecosystems.  In:  Schmidt,  W3anan  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource;  1989  October  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  62-71. 

Hegyi,  F.  1974.  A  simulation  model  for  managing  jack- 
pine  stands.  In:  Fries,  J.,  ed.  Growth  models  for  tree 
and  stand  simulation.  Res.  Notes  30.  Stockholm:  Royal 
College  of  Forestry:  74-90. 

Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  black  and  grizzly 
bears  in  the  Yellowstone  area.  In:  Bears — their  biology 
and  management;  International  Conference  on  Bear 
Research  and  Management.  Madison,  WI:  International 
Bears  Association.  5:  166-173. 

Kendall,  K.  C;  Arno,  S.  F.  1990.  Whitebark  pine— an 
important  but  endangered  wildlife  resource.  In: 
Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposiima  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain 
resource;  1989  October  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
264-273. 

Kipfer,  T.  R.  1992.  Post-logging  stand  characteristics  and 
crown  development  of  whitebark  pine  (Pinus  albicau- 
lis).  Bozeman,  MT:  Montana  State  University,  Depart- 
ment of  Earth  Science.  83  p.  Thesis. 

Lorimer,  C.  G.  1983.  Tests  of  age-independent  competition 
indices  for  individual  trees  in  natural  hardwood  stands. 
Forest  Ecology  and  Management.  6:  343-360. 

Marston,  R.  A.;  Anderson,  J.  E.  1991.  Watersheds  and 
vegetation  of  the  Greater  Yellowstone  Ecosystem.  Con- 
servation Biology.  5(3):  338-346. 

McCaughey,  W.  W.;  Schmidt,  W.  C.  1990.  Autecology  of 
whitebark  pine.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposivun  on 
whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource;  1989  October  29-31; 


134 


Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  85-96. 

Powers,  T.  M.  1991.  Ecosystem  preservation  and  the 
economy  in  the  Greater  Yellowstone  Area.  Conservation 
Biology.  5(3):  395-404. 

Schmidt,  W.  C.;  McCaughey,  W.  W.  1990.  Whitebark 
pine — a  subalpine  species  needing  silvicultural  atten- 
tion. In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  October  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
tiu-e,  Forest  Service,  Intermountain  Research  Station: 
373-374. 

Spurr,  S.  H.;  Barnes,  B.  V.  1980.  Forest  ecology.  3d  ed. 
New  York:  John  Wiley  and  Sons.  687  p. 

Tome,  M.;  Burkhart,  H.  E.  1989.  Distance-dependent  com- 
petition measures  for  predicting  growth  of  individual 
trees.  Forest  Science.  35(3):  816-831. 

Wagner,  R.  G.;  Radosevich,  S.  R.  1991a.  Neighborhood 
predictors  of  interspecific  competition  in  young 


Douglas-fir  plantations.  Canadian  Journal  of  Forest  Re- 
search. 21:  821-828. 
Wagner,  R.  G.;  Radosevich,  S.  R.  1991b.  Interspecific  com- 
petition indices  and  other  factors  influencing  the  perfor- 
mance of  Douglas-fir  saplings  in  the  Oregon  Coast 
Range.  Canadian  Journal  of  Forest  Research.  21:  829- 
835. 

Weaver,  T.;  Forcella,  F.;  Dale,  D.  1986.  Cone  production 
in  Pinus  albicaulis  forests.  In:  Shearer,  R.  C,  comp. 
Proceedings — conifer  tree  seed  in  the  inland  mountain 
west  symposium;  1985  August  5-6;  Missoula,  MT.  Gen. 
Tech.  Rep.  INT-203.  Ogden,  UT:  U.S.  Department  of 
Agriciilture,  Forest  Service,  Intermotmtain  Research 
Station:  68-76. 

Weaver,  T.;  Forcella,  F.;  Dale,  D.  1990.  Stand  develop- 
ment in  whitebark  pine  woodlands.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposimn  on  whitebark  pine  ecosystems:  ecology 
and  management  of  a  high-mountain  resource;  1989 
October  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agricvilture,  Forest  Ser- 
vice, Intermountain  Research  Station:  151-155. 


135 


FIRE  ECOLOGY  OF  WHITEBARK  PINE 
FORESTS  OF  THE  NORTHERN  ROCKY 
MOUNTAINS,  U.SA. 

Penelope  Morgan 
Stephen  C.  Bunting 
Robert  E.  Keane 
Stephen  F.  Amo 


Abstract— Fires  once  occurred  at  intervals  between  30  and  300 
years  in  whitebark  pine  (Pinus  albicaulis)  forests  in  the  Northern 
Rocky  Mountains,  U.S.A.,  but  since  the  early  1900's  fewer  fires 
have  occurred,  contributing  to  declining  abundance  of  whitebark 
pine.  In  the  absence  of  fire  or  other  major  disturbance,  whitebark 
pine  is  replaced  by  other  conifers  on  most  of  the  upper  subalpine 
landscape.  Whitebark  pines  often  survive  low-intensity  surface 
fires.  Large  stand-replacement  fires  also  benefit  this  species  by 
creating  the  open,  burned  sites  where  regeneration  is  most 
successful. 


In  the  Northern  Rocky  Mountains  of  western  North 
America,  whitebark  pine  (Pinus  albicaulis)  historically 
dominated  many  upper  subalpine  forests.  These  high- 
elevation  forests  usually  have  poorly  developed,  rocky 
soils  and  are  often  located  within  wilderness  or  roadless 
areas.  As  a  consequence,  whitebark  pine  is  seldom  har- 
vested for  forest  products,  but  it  is  important  for  scenic, 
watershed,  and  wildlife  habitat  values. 

Whitebark  pine  dominates  middle-  and  late-successional 
stages.  In  the  absence  of  major  disturbance,  however, 
whitebark  pine  is  eventually  replaced  by  the  more  shade- 
tolerant  subalpine  fir  (Abies  lasiocarpa)  and  Engelmann 
spruce  (Picea  engelmannii)  in  most  of  its  range  in  the  upper 
subalpine  forest  zone  of  the  Northern  Rocky  Mountains 
(fig.  1).  Whitebark  pine  is  a  common  serai  component 
of  upper  subalpine  forests  found  on  the  Abies  lasiocarpa/ 
Vaccinium  scoparium,  A.  lasiocarpa/Luzula  hitchcockii, 
and  A.  lasiocarpa/Arnica  cordifolia  habitat  types  (Pfister 
and  others  1977;  Steele  and  others  1981,  1983;  Weaver 
and  Dale  1974).  These  sites  are  cold,  with  July  mean 
temperatures  averaging  13  to  15  °C  (Pfister  and  others 
1977).  Although  annual  precipitation  averages  610  to 
1,780  mm,  summer  drought  is  common  (Arno  and  Hoff 
1989;  Pfister  and  others  1977;  Weaver  and  Dale  1974). 


Whitebark  pine  is  also  found  in  pure  stands  on  relatively 
dry  and  severe,  windswept  sites  near  timberline  where  it 
is  the  climax  tree  species  (Amo  and  HofF  1989).  It  is  the 
sole  climax  tree  species  on  Pinus  albicaulis  habitat  types 
in  Montana,  central  Idaho,  and  western  Wyoming,  and  in 
southern  Canada  (Amo  and  Hoff  1989;  Steele  and  others 
1981, 1983).  Whitebark  pine  and  subalpine  fir  are  climax 
codominants  on  the  Pinus  albicaulis-Abies  lasiocarpa  habi- 
tat types  where  subalpine  fir  growth  is  stunted  in  the  severe 
microclimate  (Pfister  and  others  1977).  Climax  whitebark 
pine  forests  are  usually  open,  with  smcdl  patches  of  trees 
of  mixed  ages  interspersed  with  meadows  (fig.  2).  Average 
July  mean  temperatures  are  10  to  12  °C  with  severe  sum- 
mer droughts  and  frosts;  the  annual  precipitation  of  71  to 
153  cm  falls  mostly  as  snow  (Arno  and  Hoff  1989;  Pfister 
and  others  1977). 

Fires  are  very  important  to  regeneration  and  survival  of 
whitebark  pine  on  sites  where  it  is  serai.  Whitebark  pine 
often  survives  these  low-intensity  surface  fires,  which  more 
easily  kill  associated  conifers  (fig.  3).  Stand-replacing  fires 
also  benefit  whitebark  pine,  for  although  all  trees  are  usu- 
ally killed,  whitebark  pine  regenerates  on  burned  sites  more 
successfully  than  many  associated  tree  species  (Tomback 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Penelope  Morgan  is  Associate  Professor,  Department  of  Forest  Re- 
sources, University  of  Idaho,  Moscow,  ID  83843;  Stephen  C.  Bunting  is 
Professor,  Department  of  Range  Resources,  University  of  Idaho,  Moscow, 
ID  83843;  Robert  E.  Keane  and  Stephen  F.  Amo  are  Research  Ecologist 
and  Research  Forester,  Intermountain  Fire  Sciences  Laboratory,  Inter- 
mountain  Research  Station,  Forest  Service,  U.S.  Department  of  Agricul- 
ture, P.O.  Box  8089,  Missoula,  MT  59801. 


Figure  1 — Subalpine  fir  and  Engelmann  spruce 
now  dominate  many  stands  and  whole  landscapes 
where  whitebark  pine  was  historically  abundant. 
Whitebark  pine  is  declining  in  abundance  even 
where  blister  rust  is  uncommon,  such  as  on  this 
site  east  of  Yellowstone  National  Park  in  Wyoming. 


136 


and  others  1990)  (fig.  4).  Large  stand-replacing  fires  are 
infi-equent,  usually  occurring  only  during  windy  conditions 
after  prolonged  drought.  The  fires  that  burned  in  and 
around  Yellowstone  National  Park  in  1988  were  spectacu- 
lar examples  of  the  large,  high-intensity  fires  that  periodi- 
cally burn  within  whitebark  pine  and  adjacent  forest  types. 
These  fires  burned  whitebark  pine  habitats  in  a  patchy, 
stand-replacing  manner. 

During  the  10,000  years  since  development  of  forests  af- 
ter the  last  glacial  retreat,  fires  have  had  a  major  influence 
on  the  structure  and  composition  of  forests  in  the  Northern 
Rocky  Mountains  (Arno  1980).  Fire  occurrence  has  been 
significantly  altered  by  human  activity.  Native  Americans 
used  fire  to  manipulate  vegetation,  for  hunting,  for  commu- 
nication, and  for  other  purposes  (Pyne  1982).  More  recently, 
humans  have  affected  fire  occiurence  by  purposefully  or  in- 
advertently igniting  fires,  through  fire  suppression,  and  by 
grazing  domestic  livestock,  logging,  or  otherwise  altering 
the  fuels  available  to  burn.  Efi"orts  to  suppress  fires  have 
become  increasingly  effective  since  about  1935  (Arno  1980; 
Pyne  1982). 


Figure  3 — Whitebark  pine  trees  often  survive  fires  and  individual  trees  may  survive  multiple 
fires.  Information  on  fire  frequency  and  effects  is  derived  from  dates  of  fire  scars  on  tree 
sections  such  as  this  one  from  a  208-year-old  whitebark  pine  tree  in  the  Shoshone  National 
Forest  in  northwestern  Wyoming.  First  scarred  by  fires  when  only  12.2  cm  in  diameter  at 
the  base,  it  was  again  scarred  60  years  later.  It  was  cut  in  1 988, 1 1 0  years  later. 


Figure  2 — Whitebark  pine  often  is  the  only  tree 
that  can  grow  on  harsh  sites.  Here,  in  the  Challis 
National  Forest  in  central  Idaho,  whitebark  pine  is 
the  climax  tree  species. 


137 


Figure  4 — Young  whitebark  pine  trees  regener- 
ated successfully  following  a  stand-replacement 
fire  in  northwestern  Wyoming.  All  trees  were  killed 
by  the  fire,  which  occurred  55  years  before  this 
photograph  was  taken  in  1 988. 


WHITEBARK  PINE  DECLINE 

Whitebark  pine  has  decHned  in  abundance  in  major  por- 
tions of  its  range  (Arno  1986).  Once  important  on  10  to  15 
percent  of  the  forested  landscape  in  the  Northern  Rocky 
Mountains  (Arno  and  Hoff  1989),  whitebark  pine  mortality 
rates  averaged  42  percent  over  the  last  20  years  in  western 
Montana  (Keane  and  Arno  1993).  Arno  and  others  (1993) 
found  that  for  a  200-ha  study  area  in  Montana,  the  per- 
centage of  stands  with  at  least  20  percent  basal  area  of 
mature  whitebark  pine  declined  from  37  percent  in  1900 
to  20  percent  in  1991;  14  percent  of  the  area  was  dominated 
by  whitebark  pine  in  1900;  none  was  so  dominated  in  1991. 

Whitebark  pine  decline  is  most  pronounced  on  the  more 
productive  sites  where  subalpine  fir  and  Engelmann  spruce 
are  highly  competitive  (Arno  1986;  Ciesla  and  Fumiss  1986; 
Keane  and  Arno  1993;  Kendall  and  Arno  1990).  Declining 
whitebark  pine  threatens  wildlife  habitat  because  the  seeds 
of  whitebark  pine  are  a  valuable  food  for  many  birds  and 
small  mammals,  including  the  endangered  grizzly  bear 
(Kendall  and  Arno  1990). 

Recent  decline  in  whitebark  pine  abundance  is  linked 
to  less  frequent  fires  (Keane  and  Morgan,  these  proceed- 
ings; Keane  and  others  1990).  Fires  in  whitebark  pine  for- 
ests occurred  at  mean  intervals  of  30  to  300  years  based  on 
fire  history  information  derived  from  fire  scars  and  stand 
ages  (table  1).  Fewer  fires  in  the  last  50  to  100  years  (Arno 
and  Hoff  1989;  Morgan  and  Bunting  1990)  have  resulted 
in  extensive  changes  in  the  composition  of  forests  in  the 
high-elevation  landscapes  of  the  Northern  Rocky  Mountains, 
Composition  of  subalpine  forests  has  shifted  dramatically 
toward  dominance  by  subalpine  fir  and  Engelmann  spruce 
(Keane  and  others  1993).  The  decline  of  whitebark  pine 
has  been  further  exacerbated  by  the  introduced  blister  rust 
{Cronartium  ribicola)  and  the  native  mountain  pine  beetle 
(Dendroctonus  ponderosae),  both  of  which  kill  whitebark 
pine  but  not  subalpine  fir  or  Engelmann  spruce. 


STAND-REPLACING  FIRES 

In  conditions  of  extreme  drought  lasting  more  than 
2  years,  fires  ignited  by  lightning  and  fanned  by  high  winds 
can  rapidly  spread  and  kill  trees  in  large  patches.  These 
fires  usually  bum  in  other  forest  types  as  well,  converting 
large  segments  of  the  landscape  to  early  successional  plant 
communities.  Fires  that  spread  through  forests  at  lower 
elevations  historically  burned  into  the  adjacent  whitebark 
pine  forests  (Arno  and  Hoff  1989).  In  whitebark  pine  for- 
ests, stand-replacing  fires  typically  spread  on  the  ground 
(Lasko  1990).  Fires  may  kill  trees  by  scorching  foliage  or 
by  heating  the  bole  or  roots  to  lethal  temperatures.  Some- 
times, crown  fires  occiu-  that  bum  through  the  tree  crowns, 
killing  all  trees  in  their  paths  (Lasko  1990). 

Stand-replacing  fires  provide  important  opportunities 
for  whitebark  pine  to  regenerate.  Many  competing  tree 
species  rely  on  the  wind  to  disseminate  seed.  Whitebark 
pine  has  a  distinct  advantage  in  regenerating  following 
extensive  disturbances  (Tomback  and  others  1990).  The 
Clark's  nutcracker  {Nucifraga  columbiana)  conmionly  trans- 
ports seeds  several  kilometers  (Hutchins  and  Lanner  1982). 
These  birds  prefer  open,  burned  areas  for  caching  seeds 
(Tomback  and  others  1990).  Thus,  although  large  fires  are 
infrequent,  they  are  ecologically  important  in  maintaining 
extensive  whitebark  pine  forests  on  the  landscape. 

LOW-INTENSITY  SURFACE  FIRES 

Low-intensity  surface  fires  also  influence  the  relative 
abundance  of  whitebark  pine  on  the  landscape.  Such 
fires  are  more  frequent  and  smaller  in  extent  than  stand- 
replacing  fires.  Low-intensity  fires  generally  kill  young 
whitebark  pine  and  both  large  and  small  subalpine  fir. 
Such  fires  can  result  in  open,  parklike  stands  of  nearly 
pure  whitebark  pine  (Arno  1986).  Some  fires  probably 
burned  as  low-intensity  surface  fires  but  later  became 
stand-replacing  fires  when  burning  conditions  were  more 
severe.  That  whitebark  pine  trees  often  survive  surface 
fires  is  evidenced  by  the  many  living  trees  that  have  scars 
from  one  or  more  fires  (fig.  1). 


Table  1— Fire  frequency  from  whitebark  pine  forests  expressed 
as  the  mean  and  range  (in  parentheses)  of  the  years 
between  fires 


Fire  frequency 

Geographical  area  and  reference 

144  (55  to  304) 

Bob  Marshall  Wilderness  Complex,  north- 

western fy/lontana  (Keane  and  others  1 993) 

80  (50  to  300) 

Bitterroot  Mountains,  Montana  (Arno  1980) 

30  to  41  (4  to  78) 

1 00-  to  300-ha  stands  where  subalpine  fir 

is  climax,  Montana  (Arno  1986) 

29  (13  to  46) 

10  stands  within  100  ha,  northwestern 

Wyoming  (Morgan  and  Bunting  1990) 

300 

Lodgepole  pine  forests  adjacent  to  but 

at  lower  elevations  than  whitebark  pine. 

Yellowstone  National  Park  (Romme  1 982) 

138 


These  low-intensity  fires  are  more  common  on  relatively 
dry  sites,  occurring  only  where  stand  structures,  fuel  ac- 
cumulation, and  microclimatic  conditions  are  conducive. 
Thus,  such  fires  result  in  many  small  burned  patches,  in- 
creasing landscape  heterogeneity. 

Where  fires  are  more  fi'equent,  they  are  more  likely  to  be 
of  low  intensity.  Morgan  and  Bimting  (1990)  documented 
very  fi'equent  low-intensity  fires  on  a  relatively  dry  site 
supporting  serai  whitebark  pine  in  open,  parkHke  stands 
within  a  100-ha  area  in  northwestern  Wyoming.  There  the 
mean  interval  between  fires  was  33  years  prior  to  1867. 
Fires  were  much  more  common  prior  to  1850  than  they 
have  been  since  then  (Morgan  and  Bunting  1990). 

FIRE  REGIMES 

Through  time,  most  whitebark  pine  forests  experience  a 
mixture  of  stand-replacement  and  low-intensity  fires.  The 
frequency  of  these  types  of  fires  within  a  given  landscape 
will  vary  with  landscape  complexity  and  heterogeneity. 
Stand-replacing  fires  are  more  common  diiring  regional 
droughts,  and  often  burn  large  patches  regardless  of  fuel 
loading  or  stand  condition.  Fire  behavior  and  effects  are 
also  influenced  by  the  stand  structure  and  fuel  accumula- 
tion, which  are  in  part  determined  by  the  time  since  last 
bum. 

Where  whitebark  pine  is  cHmax,  fires  are  infrequent  and 
generally  of  low  intensity.  In  whitebark  pine  krummholz 
and  ribbon  forests,  fires  are  infrequent  and  of  variable  in- 
tensity. When  fires  do  occ\ir,  many  trees  die  and  regenera- 
tion is  very  slow.  Keane  and  others  (1990)  predict  where 
blister  rust  infection  rates  are  high,  climax  whitebark  pine 
forests  will  convert  to  herbaceous  or  shrub  commtmities 
following  fire. 

Stand-replacing  fires  are  more  common  where  whitebark 
pine  is  a  serai  dominant.  Stand-replacing  fires  become  in- 
creasingly likely  with  advancing  succession  (Fischer  and 
Clayton  1983;  Morgan  and  Bimting  1990). 

SUCCESSION  FOLLOWING  FIRE 

Fire  is  a  key  process  affecting  serai  whitebark  pine  forest 
structure  sind  composition.  Successional  patterns  on  sites 
where  whitebark  pine  is  serai  are  predictable  (fig.  5),  but 
they  are  not  closely  tied  to  stand  age  or  time  since  last  dis- 
tiu-bance  (Mattson  and  Reinhart  1990).  The  stand  struc- 
ture and  the  microsites  created  vary  from  fire  to  fire.  Con- 
ditions for  successful  regeneration  of  tree  seedlings  are 
sporadic,  depending  on  favorable  climatic  and  site  condi- 
tions. Althoiigh  both  subalpine  fir  and  whitebark  pine 
may  establish  soon  after  a  fire,  it  may  take  a  half-century 
or  longer  for  a  forest  to  develop. 

Whitebark  pine  is  one  of  the  first  tree  species  to  become 
established  in  abimdance  following  stand-replacing  fires 
(Weaver  and  Dale  1974).  As  a  consequence,  it  often  domi- 
nates initially,  often  for  up  to  225  years  or  more  (Loope 
and  Gruell  1973;  Morgan  and  Bimting  1990).  Early  serai 
stands  are  dominated  by  whitebark  pine  seedlings  and  sap- 
lings growing  along  with  a  dense  herbaceous  and  shrub 
understory.  Subalpine  fir  seedlings  are  often  present, 
especially  close  to  parent  trees  that  survived  the  fire,  but 
they  grow  more  slowly  than  whitebark  pine  trees  (Arno 


Late  Serai 
(ABLA,  PEN, 
PIAL,  PICO) 


Shrub/herb 


Late  Mid-Seral 
(ABLA,  PIAL, 
PICO,  PEN) 


Early  Sera! 
(PIAL,  PICO 
ABLA,  PEN) 


Mid-Serai 
(PIAL,  PICO, 
ABLA,  PEN) 


^^^^^^ 


Stand-replacing  fire 
Return  to  shrub/herb  stage 


Low-intensity  fire 
Many  PIAL  survive 


Figure  5 — Generalized  forest  succession  following 
fires  on  sites  where  whitebark  pine  (PIAL)  is  re- 
placed by  subalpine  fir  (ABLA)  and  Engelmann 
spruce  (PEN)  with  advancing  succession.  Lodge- 
pole  pine  (PICO)  is  a  common  associate.  Low- 
intensity  surface  fires  may  occur  at  any  stage 
but  are  most  likely  in  mid-seral  stands.  Stand- 
replacing  fires  are  the  norm  in  late-seral  stands. 
Tree  species  listed  in  order  of  abundance. 
Adapted  from  Fischer  and  Clayton  (1983). 


and  Hoff  1989).  Whitebark  pine  seedlings  are  more  abun- 
dant than  subalpine  fir  seedlings  in  large  burned  areas 
because  whitebark  pine  seedlings  are  dispersed  farther  by 
the  Clark's  nutcracker.  Subalpine  fir  seeds  are  dispersed 
by  the  wind.  Many  standing  snags  and  fallen  logs  are 
present.  Most  whitebark  pine  trees  do  not  produce  large 
numbers  of  seeds  imtil  at  least  age  70;  most  do  not  produce 
any  cones  xmtil  age  50  (Morgan  and  Bvmting  1992).  With 
time,  whitebark  pine  is  gradually  replaced  by  subalpine  fir 
and  Engelmann  spruce.  In  late-seral  stands,  both  the  tree 
canopy  and  the  imderstory  are  dominated  by  many  subal- 
pine fir  trees. 

MANAGEMENT  IMPLICATIONS 

Continued  decline  in  whitebark  pine  abundance  threat- 
ens to  dramatically  reduce  the  availability  of  seeds  for  the 
many  animals  that  rely  on  them  as  a  food  source  (Arno 
1986). 

Fire  exclusion  greatly  reduces  opportimities  for  regen- 
eration of  whitebark  pine.  Cone  production  is  higher  in 
the  stands  where  whitebark  pine  is  healthy  and  dominant 
(Morgan  and  Bunting  1992).  With  blister  rust  reducing 
cone  production  and  killing  parent  trees  (Arno  and  Hoff 
1989),  the  seed  available  for  tree  regeneration  is  rapidly 


139 


declining  (Amo  1986;  Keane  and  Amo  1993).  As  cone  pro- 
duction declines,  animals  eat  more  of  the  seeds,  leaving 
fewer  to  regenerate.  Whitebark  pine  cone  production  de- 
clines with  advancing  succession  (Morgan  and  Bunting 
1992)  and  as  infection  by  blister  rust  increases  (Arno  and 
Hoff  1989).  Given  the  rapidly  declining  abundance  of  white- 
bark  pine  in  some  regions,  we  must  act  quickly  to  create  op- 
portunities for  individuals  that  are  resistant  to  blister  rust 
to  regenerate.  Although  whitebark  pine  is  very  sensitive 
to  blister  rust,  some  individual  trees  are  genetically  resis- 
tant to  infection  (Hoff  and  others  1990).  If  we  do  not  create 
opportunities  for  those  trees  to  regenerate  before  their  cone 
production  declines  or  they  die  in  advancing  succession, 
opportunities  for  enhancing  natural  mechanisms  of  white- 
bark pine  recovery  will  be  lost. 

Three  options  are  available  for  improving  the  health  and 
productivity  of  whitebark  pine  stands  that  are  now  domi- 
nated by  subalpine  fir  and  spruce.  One  option  is  creating 
forest  openings  through  timber  harvest,  girdling,  or  other- 
wise killing  trees  mechanically.  Cutting  trees  to  create 
openings  to  encourage  regeneration  of  whitebark  pine  is 
possible  but  may  not  be  economically  feasible.  Although 
there  are  often  small  whitebark  pine  trees  growing  in  the 
imderstory  of  mixed  conifer  stands,  they  are  often  very  old 
and  of  poor  vigor  and  are  therefore  unlikely  to  respond 
when  larger  trees  are  removed. 

Another  option  is  the  liberal  use  of  prescribed  fire  to  cre- 
ate openings  for  regeneration  and  to  favor  whitebark  pine 
in  stands  now  codominated  by  subalpine  fir.  Managers  can 
purposefully  ignite  fires  or  allow  lightning  fires  to  burn 
under  carefully  prescribed  conditions  of  weather,  fuel,  and 
location.  Historically,  fire  was  the  primary  natural  distur- 
bance, and  it  may  be  the  most  practical  tool  for  managing 
whitebark  pine  considering  economics,  policy,  and  topo- 
graphic limitations  for  using  timber  harvest. 

A  third  option,  to  use  a  combination  of  techniques,  may 
be  most  successful.  Arno  and  Keane  are  involved  in  an  ef- 
fort to  test  alternative  techniques  for  perpetuating  white- 
bark pine  in  a  blister  rust-infected  area  of  the  Bitterroot 
National  Forest  in  Montana. 

ACKNOWLEDGMENTS 

Research  funding  was  provided  through  the  USDA  Forest 
Service,  Intermountain  Research  Station.  This  is  Contribu- 
tion No.  693  of  the  Idaho  Forest,  Wildlife  and  Range  Experi- 
ment Station,  University  of  Idaho,  Moscow,  ID  83843. 
We  thank  Bill  Fischer  and  Elizabeth  Reinhardt  for  their 
comments. 

REFERENCES 

Arno,  S.  1980.  Forest  fire  history  in  the  Northern  Rockies. 
Journal  of  Forestry.  78(8):  460-465. 

Arno,  S.  1986.  Whitebark  pine  cone  crops — a  diminishing 
source  of  wildlife  food?  Western  Journal  of  Applied  For- 
estry. 1(3):  92-94. 

Arno,  S.  F.;  Hoff,  R.  J.  1989.  Silvics  of  whitebark  pine 
{Pinus  albicaulis).  Gen.  Tech.  Rep.  INT-253.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station,  lip. 


Arno,  S.  F.;  Reinhardt,  E.  D.;  Scott,  J.  H.  1993.  Forest 
structure  and  landscape  patterns  in  the  subalpine  lodge- 
pole  pine  t)^es:  a  procedure  for  quantifying  past  and 
present  conditions.  Gen.  Tech.  Rep.  INT-294.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service, 
Intermountain  Research  Station.  17  p. 

Ciesla,  W.  M.;  Furniss,  M.  M.  1986.  Idaho's  haimted  forests. 
American  Forests.  81(8):  23-25. 

Fischer,  W.  C;  Clayton,  B.  D.  1983.  Fire  ecology  of  Montana 
forest  habitat  types  east  of  the  Continental  Divide.  Gen. 
Tech.  Rep.  INT-141.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermountain  Forest  and 
Range  Experiment  Station.  83  p. 

Hoff,  R.  J.;  Bingham,  R.  T.;  McDonald,  G.  I.  1980.  Relative 
blister  rust  resistance  of  white  pines.  European  Journal 
of  Forest  Pathology.  10:  307-316. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Keane,  R.  E.;  Arno,  S.  F.  1993.  Rapid  decline  of  whitebark 
pine  in  western  Montana:  evidence  from  10-year  meas- 
urements. Western  Journal  of  Applied  Forestry.  8(2): 
44-47. 

Keane,  R.  E.;  Arno,  S.  F.;  Brown,  J.  K;  Tomback,  D.  F. 
1990.  Modelling  stand  dynamics  in  whitebark  pine 
forests.  Ecological  Modelling.  51:  73-95. 

Keane,  R.  E.;  Morgan,  P.;  Menakis,  J.  1993.  Landscape 
assessment  of  the  decline  of  Pinus  albicaulis  in  the 
Bob  Marshall  Wilderness  Complex,  Montana,  USA. 
Missoula,  MT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station,  Intermountain 
Fire  Sciences  Laboratory.  Review  draft. 

Kendall,  K  C;  Arno,  S.  F.  1990.  Whitebark  pine — an  impor- 
tant but  endangered  wildlife  resource.  In:  Schmidt,  W.  C; 
McDonald,  K.  J.,  comps.  Proceedings — sjmaposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station:  264-273. 

Lasko,  Richard  J.  1990.  Fire  behavior  characteristics  and 
management  implications  in  whitebark  pine  ecosystems. 
In:  Schmidt,  W.  C.;  McDonald,  K.  J.,  comps.  Proceedings — 
sjmiposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station:  319-323. 

Loope,  L.  L.;  Gruell,  G.  E.  1973.  The  ecological  role  of  fire 
in  the  Jackson  Hole  area,  northwestern  Wyoming.  Qua- 
ternary Research.  3:  425-443. 

Mattson,  D.  J.;  Reinhart,  D.  P.  1990.  Whitebark  pine  on  the 
Mount  Washburn  Massif,  Yellowstone  National  Park.  In: 
Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  Proceedings- 
symposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station:  106-117. 

Morgan,  P.;  Bunting,  S.  C.  1990.  Fire  effects  in  whitebark 
pine  forests.  In:  Schmidt,  W.  C;  McDonald,  K.  J.,  comps. 


140 


Proceedings — symposium  on  whitebark  pine  ecosystems: 
ecology  and  management  of  a  high-momitain  resource; 
1989  March  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep. 
INT-270.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Research  Station:  166-170. 

Morgan,  P.;  Bunting,  S.  C.  1992.  Cone  crops  in  whitebark 
pine  (Pinus  albicaulis)  forests  in  the  Greater  Yellowstone 
Area.  Moscow,  ID:  University  of  Idaho,  College  of  For- 
estry, Wildlife  and  Range  Sciences.  21  p.  Review  draft. 

Pfister,  R.  D.;  Kovalchik,  B.  L.;  Arno,  S.  F.;  Presby,  R.  C. 
1977.  Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Experi- 
ment Station.  174  p. 

Pyne,  S.  J.  1982.  Fire  in  America.  Princeton,  NJ:  Princeton 
University  Press.  654  p. 

Romme,  W.  H.  1982.  Fire  and  landscape  diversity  in  subal- 
pine  forests  of  Yellowstone  National  Park.  Ecological 
Monographs.  52:  199-221. 

Steele,  R.;  Pfister,  R.  D.;  Ryker,  R.  A.;  Kittams,  J.  A.  1981. 
Forest  habitat  types  of  central  Idaho.  Gen.  Tech.  Rep. 


INT-114.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  138  p. 

Steele,  R.;  Cooper,  S.  V.;  Ondov,  D.  M.;  Roberts,  D.; 

Pfister,  R.  D.  1983.  Forest  habitat  types  of  eastern  Idaho- 
western  Wyoming.  Gen.  Tech.  Rep.  INT- 144.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Forest  and  Range  Experiment  Station.  122  p. 

Tomback,  D.  F.;  Hoffman,  L.  A.;  Sund,  S.  K.  1990.  Coevo- 
lution  of  whitebark  pine  and  nutcrackers:  implications 
for  forest  regeneration.  In:  Schmidt,  W.  C;  McDonald, 
K.  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  118-130. 

Weaver,  T.;  Dale,  D.  1974.  Pinus  albicaulis  in  central 
Montana:  environment,  vegetation  and  production. 
American  Midland  NaturaHst.  92:  222-230. 


141 


VEGETATION  DISTRIBUTION  AND 
PRODUCTION  IN  ROCKY  MOUNTAIN 
CLIMATES— WITH  EMPHASIS  ON 
WHITEBARK  PINE 

Tad  Weaver 


Abstract — The  distribution  and  production  of  vegetation  on  the 
altitudinal  gradient  (grassland-forest-alpine)  was  plotted  against 
climatic  parameters  to  evaluate  hypothetical  controlling  factors. 
(1)  Whitebark  pine  (Pinus  albicaulis)  is  likely  excluded  from 
higher  zones  by  a  cool  growing  season  or  wind-induced  drought. 
It  is  probably  not  excluded  by  low  temperatures  occurring  during 
its  hardening,  hard,  or  dehardening  seasons.  (2)  While  the  lower 
physiological  limit  of  whitebark  pine  is  probably  set  by  drought, 
its  lower  realized  limit  is  directly  set  by  subalpine  fir  (Abies  lasio- 
carpa)  and  lodgepole  pine  (Pinus  contorta)  competitors  and  indi- 
rectly set  by  factors  that  control  their  distribution.  (3)  The  upper 
limits  for  most  other  dominant  species  are  probably  set  by  grow- 
ing season  temperature.  The  lower  Umits  are  likely  set  by  com- 
petition down  to  the  cedar-hemlock  (Thuja  plicata  I  Tsuga  hetero- 
phylld)  zone  and  by  drought  in  drier  areas.  (4)  Production  is 
strongly  correlated  (r^  =  0.86)  with  growing  season  length  (soil 
thawed  season  minus  dry  soil  days).  Multiplying  season  length 
by  average  temperature  did  not  improve  the  growing  season  pre- 
dictor, perhaps  because  vegetation  at  each  altitude  is  especially 
adapted  to  temperatures  in  its  zone. 


Vegetation  composition  and  structure  vary  along  gradi- 
ents of  temperature  and  precipitation  v^^hether  the  condi- 
tion changes  with  altitude  (Daubenmire  1956;  Gams  1931; 
Weaver  1980)  or  geography  (Holdridge  1967;  Walter  1973; 
Whittaker  1975).  Graphical  devices  based  on  simple  pa- 
rameters such  as  average  annual  temperature  and  total 
annual  precipitation  (Holdridge  1967;  Whittaker  1975) 
are  workable  predictors  of  the  vegetation  growing  in  par- 
ticular climates.  It  seems  obvious,  however,  that  these 
devices  succeed,  not  because  the  climatic  parameters  used 
are  causal,  but  because  the  parameters  are  correlated 
with  causal  climatic  factors. 

Production  of  vegetation  is  largely  determined  by  cli- 
mate and  is  therefore  expected  to  be  highest  where  tem- 
perature, moisture,  and  nutrients  are  simidtaneously 
favorable.  In  Rocky  Mountain  vegetation  one  therefore 
expects  production  to  be  highest  in  low-altitude  (warm) 
moist  forests  (Thuja-Tsuga)  and  to  decline  both  with  in- 
creasing altitude  (because  of  cooling  and  reduced  nutri- 
ent availability)  (Weaver  1979)  and  with  decreasing  alti- 
tude (because  of  decreasing  water  availability).  While  it  is 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Tad  Weaver  is  Plant  Ecologist,  Biology  Department,  Montana  State 
University,  Bozeman,  MT  59717. 


rarely  done,  vegetation  production  should  also  be  success- 
fully plotted  in  a  Hutchinsonian  (1958)  hyperspace  with 
axes  that  are  either  physiologically  meaningful  or  surro- 
gates well  correlated  with  physiologically  meaningful  axes. 

The  object  of  this  paper  is  to  correlate  vegetation 
performance — survival  and  production — ^with  physiologi- 
cally meaningftd  aspects  of  temperature  and  moisture. 
Study  of  the  distribution  of  species  on  these  prestunptively 
causal  axes  will  eliminate  some  hypotheses  of  cause  and 
sharpen  others  for  experimental  tests.  While  whitebark 
pine  (Pinus  albicaulis)  is  the  primary  subject  of  this  paper, 
the  approach  could  be  applied  to  other  species,  as  well. 

METHODS 

Impacts  of  climatic  factors  on  vegetation  distribution 
were  evaluated  by  comparing  factor  levels  among  seg- 
ments on  a  vegetational  gradient.  In  essence,  the  method 
allows  one  to  deduce  (1)  that  a  presmnptive  factor  that 
does  not  vary  between  vegetation  t3T)es  is  not  controlling 
the  vegetation  changes  observed,  and  (2)  that  a  presimip- 
tive  factor  that  does  change  between  vegetation  types  de- 
serves further  discussion  as  directly  controlhng,  indirectly 
controlling,  or  a  noncausal  correlate.  The  method  is  illus- 
trated here  with  discussion  of  whitebark  pine,  but  it  could 
be  applied  to  many  other  species  as  well. 

The  Data 

Environmental  zones  were  identified  by  climax  vegeta- 
tion occupying  them  (Daubenmire  1943;  Daubenmire  and 
Daubenmire  1968;  Huschle  and  Hironaka  1980).  They  in- 
cluded, from  low  to  high  altitude,  desert  shrub,  dry  grass- 
land, warm  forest,  cool  forest,  and  alpine  ecosystems. 
Specific  zones  are  listed  with  their  Kuchler  (1964)  type 
numbers  and  representative  stations  in  table  1. 

Climates  of  the  environmental  zones  were  characterized 
by  using  climatic  statistics  from  approximately  five  sta- 
tions in  each  environmental-vegetation  zone.  To  maintain 
the  integrity  of  the  data,  raw  data  are  plotted  wherever 
possible.  When  necessary,  medians  were  used  to  repre- 
sent "typical"  sites;  use  of  medians  deemphasizes  sites 
vrith  conditions  intermediate  between  modal  conditions 
of  adjacent  types.  The  stations  chosen  include  most  avail- 
able stations  and  may  thus  be  considered  as  a  "complete 
sample."  The  sample  may  not  be  entirely  random  because 
weather  stations  must  be  accessible;  in  higher  types,  for 
example,  the  sites  may  be  relatively  low. 


142 


Table  1 — Weather  stations  representative  of  major  Northern  Rocky  Mountain  environmental  zones. ^  Zones  are  listed  from  high  to  low  altitude 


Zone  Station  locations 


Alpine^ 

Whitebark  pine^ 
Subalpine  fir 
Douglas-fir 
Cedar-hemlock 
Ponderosa  pine 

Idaho  fescue 
Bluebunch 
wheatgrass 
Grama  grass 
Atriplex  spp. 


White  Mountain,  CA;  Niwot  Ridge,  CO 

Old  Glory,  BC;  Kings  Hill,  MT;  Ellery  Lake,  CA;  Crater  Lake,  OR 

High:  Summit,  MT;  West  Yellowstone,  MT;  Lake  Yellowstone,  MT;  Low:  Burke,  ID;  Seeley  Lake,  MT;  Hungry  Horse,  MT 
East:  Hebgen  Lake,  MT;  Lakeview,  MT;  Lamar,  WY;  Palisades  Dam,  ID;  Dixie,  ID;  West:  Libby,  MT;  Lincoln,  MT 
West  Glacier,  MT;  Sandpoint,  ID;  Pierce,  ID;  Avery,  ID;  Priest,  ID 

West:  Garden  Valley,  ID;  Poison,  MT;  Potlatch,  ID;  Kooskia,  ID;  East:  Melstone,  MT;  Busby,  MT;  Colstrip,  MT;  Lame 
Deer,  MT;  Roundup,  MT 

Gallatin  Gateway,  MT;  Virginia  City,  MT;  Mystic  Lake,  MT;  Wisdom,  MT;  Bozeman  MSU,  MT;  White  Sulphur  Springs,  MT 

Browning,  MT;  Belgrade,  MT;  Kalispell,  MT;  Ennis,  MT;  Dillon,  MT;  Augusta,  MT;  Crow  Agency,  MT;  Billings,  MT 
Ekalaka,  MT;  Jordan,  MT;  Malta,  MT;  Rock  Springs,  MT;  Rapelje,  MT;  Chester,  MT;  Great  Falls,  MT;  Fairfield,  MT 
Lovell,  WY;  Worland,  WY;  Basin,  WY;  Powell,  WY;  Deaver,  WY 


'The  vegetation  types,  with  their  Kuchler  (1964)  type  numbers  (KTXX),  are:  Deschampsia  caespitosa-Carex  (KT52),  Pinus  albicaulis  (KT15),  Abies  lasiocarpa 
(KT1 5),  Pseudotsuga  menziesii  (KT1 2),  Thuja  plicataJTsuga  heterophylla  (KT2),  Pinus  ponderosa  (KT1 1,16),  Festuca  idatioensis  (KT63),  Agropyron  spicatum 
(KT63),  Bouteloua  gracilis  (KT64),  Atripiex  (KT40).  Abies  types  from  higher  and  lower  altitudes,  Pseudotsuga  types  from  east  and  west  of  the  mountains,  and  pon- 
derosa pine  types  from  east  and  west  of  the  Rockies  are  separately  presented  to  emphasize  possible  differences. 

n"he  White  Mountain  alpine  site  (1 0,1 50  ft)  is  unusually  dry.  The  Ellery  Lake  whitebark  pine  site  has  only  precipitation  data;  temperature  data  were  therefore 
taken  from  a  similar  altitude  at  a  nearby  White  Mountain  site  (9,645  ft). 


Killing  Temperatures 

The  hypothesis  that  extreme  temperatiires  might  ex- 
clude a  plant  (for  example,  whitebark  pine)  from  an  adja- 
cent zone  was  tested  by  determining  whether  the  condi- 
tion is  more  extreme  in  the  iinoccupied  than  in  the 
occupied  zone;  if  it  is  not,  the  hypothesized  factor  is  con- 
sidered unlikely  to  be  lethal.  This  test  was  used  on  tem- 
peratures of  early  fall  frosts,  midwinter  extreme  lows, 
midwinter  average  lows,  late  fall  frosts,  midsummer  aver- 
age highs,  and  midsummer  extreme  highs.  Midwinter 
lows  £ind  midsvmimer  highs  were  tested  because  these 
temperatures  might  exceed  the  potential  tolerances  of  a 
plant;  fall  and  spring  frosts  were  considered  because  they 
might  kill  partially  hardened  plants.  Fall  frost  tempera- 
tiu-es  were  measured  as  the  absolute  low  in  the  month  in 
which  average  temperature  first  fell  below  0  °C;  where  no 
month  had  an  average  temperatvire  of  0  °C,  the  values 
were  interpolated  to  0  °C  from  adjacent  months.  Spring 
frost  temperatiu-es  were  estimated  similarly  for  the 
month  in  which  average  temperatures  first  rose  above 
0  °C.  I  argue  later  (see  growing  season)  that  average  tem- 
peratures below  0  °C  restrict  water  and  nutrient  uptake 
and  therefore  open  and  close  the  growing  season;  the  use 
of  0  °C  is  in  contrast  to  the  5  °C  used  by  many  phenolo- 
gists  (for  example,  Chang  1968).  "Absolute"  temperatures 
were  the  lowest  (or  highest)  seen  in  the  period  of  record — 
10  years  or  more. 

Starvation  Temperatiu*es 

The  plant  might  be  excluded  from  a  zone  free  from  kill- 
ing events  if  (temperatiire)  conditions  failed  to  support 
net  photosynthesis  on  an  annual  basis.  Two  indices  of 
growing  season  temperature  were  used  to  test  the  possi- 
bility of  such  starvation.  First,  temperatures  were  aver- 
aged across  all  growing  season  (defined  later)  months. 
Despite  its  common  use  in  predicting  growth  (Chang  1968; 
Larcher  1975),  this  index  is  expected  to  underestimate  the 


benefits  of  warmer  temperatvires  because  chemical  reaction 
rates  increase  exponentially  with  increasing  temperature. 
An  alternate  index  of  temperature  was  therefore  tested. 
In  the  second  index  temperatures  were  replaced  with 
growth  support  units  (gsu)  and  these  were  averaged 
across  the  season.  On  the  assimiptions  that  Q^^  =  2,  that 
very  slow  growth  begins  at  1  °C,  and  that  native  plants 
tolerate  normal  high  temperatures  of  their  zones,  the 
growth  support  unit  curve  was  constructed  by  interpolat- 
ing between  0  growth  rate  imits  (gsu)  at  0  °C,  1  gsu  at 
1  °C,  2  gsu  at  11  °C,  4  gsu  at  21  °C,  8  gsu  at  31  °C,  and 
16  gsu  at  41  °C.  That  is,  we  expect  no  growth  at  0  °C  or 
below,  1  vmit  at  1  °C,  2  imits  at  11  °C,  4  units  at  21  °C,  etc. 

Killing  Drought 

Evaporation  equals  roughly  2  mm/°C  x  month 
(Daubenmire  1956;  Nielsen  1986;  Stephenson  1990; 
Thornthwaite  1948;  Walter  1973).  As  a  result,  one  might 
index  drought  duration  by  counting  months  when  average 
precipitation  (mm/2)  is  less  than  average  temperature 
(C)  or  drought  intensity  by  summing,  across  growing  sea- 
son months,  precipitation  deficits  below  the  calculated 
balance. 

A  covmt  of  months  registering  drought  on  the  Walter  in- 
dex may  overestimate  drought  duration  where  the  deficit 
is  small  because  drought  may  open  late  or  close  early  in 
the  month.  To  minimize  this  effect  I  have  normalized 
drought  duration  from  a  scatter  diagram  of  duration 
against  deficit:  0-5  mm  =  0  months,  6-15  mm  =  1  month, 
16-35  mm  =  2  months,  36-50  mm  =  3  months,  and 
51-85  mm  =  4  months. 

I  have  used  the  simi  of  warm-season  (average  tempera- 
ture above  0  °C)  water  deficits  as  an  index  of  drought 
magnitude.  I  do  so  with  the  recognition  that,  due  to  ex- 
clusion of  important  factors  (for  example,  wind,  Penman 
1949,  and  exponential  temperature  effects,  Stephenson 
1990),  this  index  usually  underestimates  water  deficit. 


143 


Water  Starvation 


RESULTS  AND  DISCUSSION 


"Water  starvation"  should  occur  where  plants  are  not 
desiccated,  but  where  water  supplies  do  not  support  net 
photosynthesis  on  an  annual  basis.  For  example,  starva- 
tion would  occur  if  water  were  continually  available,  but 
evapotranspiration  equaled  or  exceeded  uptake  rates.  Such 
starvation  would  select  for  organisms  with  lower  ET  rates: 
Mesophytic  leaves  would  be  replaced  by  xeroph)i;ic  leaves 
and  leaf  exposiu-e  might  be  reduced  by  reduction  of  stature 
even  to  the  point  where  plants  existed  only  at  the  groimd 
surface  or  under  transparent  rock. 

Growing  Season 

For  sensitive  introduced  plants,  frost-free  season  is  of- 
ten recorded  as  an  indicator  of  season  length  (Burke  and 
others  1976;  USGS  1971). 

Since  cool  region  plants  normally  tolerate  growing  sea- 
son frosts,  I  argue  for  a  different  index.  Growth  requires 
both  leaf  activity  (photos3nithesis)  and  root  activity 
(water  and  nutrient  absorbtion).  For  plants  with  frost- 
insensitive  leaves  (Burke  and  others  1976),  midday  tem- 
peratures should  be  high  enough  for  photosynthesis  long 
before  soils  are  thawed.  Root  activity  extends  (maxi- 
mally) from  spring  soil  thaw  to  the  fall  refreeze,  so  this 
period  should  be  a  better  index  of  growing  season  length. 
Thus  I  open  and  close  the  growing  season  with  0  °C  aver- 
age monthly  air  temperatures.  So  represented,  initial  ac- 
tivity in  grasslands  begins  at  spring  soil  temperatures 
(0-25  cm)  near  0  °C  and  activity  ceases  in  the  fall  with  soil 
temperatures  as  high  as  5  °C  (Weaver,  in  preparation). 
In,  or  adjacent  to,  the  subalpine  fir  {Abies  lasiocarpa)  zone 
the  growing  season  may  exceed  this  period  because  deep 
snow  cover  may  prevent  the  freezing  of  soil  water. 

While  the  temperature-bounded  season  probably  ap- 
plies at  high-altitude  sites,  at  low-altitude  sites  soil  dry- 
ing may  close  the  growing  season  before  frost  does.  Thus, 
one  may  better  index  growing  season  as  warm  season 
months  minus  any  drought  months  included  in  the  warm 
season.  While  the  Northern  Rocky  Mountain  drought  sea- 
son comes  at  summer's  end  (Weaver  1980),  this  index  will 
also  apply  to  regions  where  the  warm  season  opens  with 
dry  months.  It  is  conceivable  that  other  periodic  factors, 
for  example  nutrients  or  pest  attack,  might  affect  growing 
season  length. 

Production 

Production  estimates  for  each  series  were  drawn  from 
the  literature  (table  2).  These  are  correlated  with  climatic 
conditions  that  might  control  them:  warm  season  length, 
warm  moist  season  length,  and  the  product  of  warm  soil 
season  length  and  temperature.  For  the  latter,  two  tem- 
perature expressions  were  tested:  average  temperature 
minus  5  °C  and  Q  (the  Q^^-based  average).  The  third 
(product)  indices  were  expected  to  be  best  because  they 
assume  that  plants  grow  when  soils  are  unfrozen-moist 
and  that  their  growth  rate  is  proportional  to  temperature 
(over  5  °C  for  the  first  linear  index  and  over  1  °C  for  the 
second  exponential  index). 


In  the  Northern  Rocky  Motmtains  vegetation  changes 
from  grass-shrub  (Atriplex,  Bouteloua,  Agropyron,  and 
Festuca)  to  conifer  {Pinus  ponderosa,  Thuja-Tsuga, 
Pseudotsuga,  and  Abies)  to  alpine  graminoid  with  increas- 
ing altitude.  Subzones  listed  parenthetically  are  described 
by  Daubenmire  (1943)  and  mapped  by  Kuchler  (1964). 
This  paper  compsu-es  weather  data  gathered  in  these  zones 
to  test  h3TJothesized  controls  of  distribution  and  production. 
The  method  is  illustrated  with  whitebark  pine. 

Upward  Limits 

I  speculate  that  the  upward  limits  of  whitebark  pine 
might  be  determined  by  lethal  factors  (winter  low  tem- 
peratures, spring-fall  fi*osts,  or  desiccation)  or  production 
deficits  (starvation)  due  to  inadequate  water  or  heat 
units.  These  h3T)otheses  are  considered  in  the  following 
paragraphs. 

If  winter  lows  are  lower  in  the  alpine  than  in  the  pine 
zone  below,  winter  cold  might  be  the  excluding  factor,  oth- 
erwise not.  Contrary  to  my  expectation,  neither  absolute 
lows  nor  average  January  minima  are  lower  in  the  alpine 
than  in  the  pine  (or  most  other)  vegetation  zone(s)  (fig.  1). 
The  failure  of  low  temperatures  to  develop  at  higher  alti- 
tudes may  be  due  to  the  high  density  of  cold  air:  Cold  air 
entering  from  the  north  stays  low  like  mercury  poured  \m- 
der  water  and  air  cooled  by  exposure  to  cold  alpine  ground 
runs  off  (Geiger  1965).  Since  extreme  temperatures  are 
not  lower  at  high  than  low  altitude,  I  deduce  that  neither 
whitebark  pine,  nor  most  other  native  dominants,  are  ex- 
cluded from  high  sites  by  extreme  low  temperatures  of  win- 
ter. An  exception  to  this  generalization  seems  to  appear  in 
high-  or  low-altitude  fi-ost  pockets,  where  accumulating 
cold  air  may  kill  trees  in  winter  or  spring  (Weaver  1990). 

For  natives  of  an  area,  I  suggest  that  frost  damage  is 
more  likely  when  plants  are  partially  hardened  (fall)  or 
partially  dehardened  (spring)  than  at  midwinter.  I  expect 
(1)  winter  to  be  delimited  approximately  by  the  fall  month 
in  which  surface  soils  fall  below  0  °C  and  the  spring 
month  in  which  they  rise  above  0  °C,  (2)  winter  to  be  a 
season  of  low  root  activity,  low  water,  and  low  nutrient 
uptake,  and  (3)  0  °C  soil  temperatures  to  be  approxi- 
mately coincident  with  0  °C  air  temperatures  (Weaver,  in 
preparation).  I  therefore  (1)  compare,  across  vegetation 
types,  observed  temperature  lows  in  these  spring-fall  sea- 
sons of  incomplete  frost  hardness  (fig.  2),  (2)  discover  no 
difference,  and  (3)  conclude  that  frosts  probably  do  not 
partition  vegetation  native  to  the  region.  The  preceding 
discussion  depends  on  the  assumption  that  plants  harden 
and  deharden  in  phase  with  air/soil  temperatures.  This 
argument  would  be  fallacious  if  day  length  were  the  pri- 
mary controller  of  the  hardening/dehardening  process. 
However,  since  both  temperature  and  correlated  day 
length  triggers  are  important  (Salisbury  and  Ross  1992), 
I  expect  frosts  to  exclude  plants  from  environmentally  dis- 
tant, but  not  environmentally  adjacent,  vegetation  types. 

While  frosts  and  winter  freezes  probably  do  not  exclude 
whitebark  pine  from  the  alpine  (or  other  plants  from  the 
vegetation  zone  immediately  above),  a  lack  of  heat  imits 


144 


Table  2 — Productivity  in  grams  per  square  meter  per  year  and  aboveground  standing  crop  (Abvegr  std  crp)  of  major  Rocky  Mountain  vegetation 
types 


Vegetation  type 

Productivity^ 

AKuonr  QtH  f^m 

g/m^/yr 

t/ha  (age.yr) 

Alpine 

135 

1.35 

Thilenlus  and  others  1974 

100-200 

1.00-2.00 

Scott  and  Billings  1 964 

Pinus  albicaulis 

'60 

^140  ^300-500^ 

VVCdVd   Cll  IVJ  L^ClIC'    1      /  H 

'25-75  (53) 

PfiQtpr  anH  nthorQ  1Q77 

r  IIOLC7I    al  IVJ  \JlllC7IO    1*7/  1 

200-700 

350  (300-500) 

Forcella  and  Weaver  1 977 

Abies  lasiocarpa 

1 00-200 

160  f300-700) 

Aoipt  and  other*;  1QfiQ 

860 

357  M06  vr) 

Whitt^^kpr  ^r\c\  Niprinn  1  07*^ 

V  V  1  II  llCir\C7l  ai  \\a  I  NIOI  ll  im   I  ^  r  w 

'95-180  (137) 

Pfister  and  others  1 977 

100-250  (Old) 

Weaver  and  Forcella  1977 

150-250 

Landis  and  Mogren  1975 

Ps6udotsuga  menziesii 

1,550 

438  (252) 

Whittakpr  and  Niprinn  1Q7^ 

'30-170  (150) 

Pfi^tpr  and  nthpr^  1Q77 

100-350  (350+) 

Weaver  and  Forcella  1 977 

Thuja  plicata- 

Tsuaa  hsteroohvlla 

870 

Hanlfiv  1  Q7fi 
rial  Moy  i  7 /  u 

550 

316  (250) 

Hanlev  1976 

1,380 

504  (103) 

Hanley  1976 

'150-330  (240) 

Pfister  and  others  1 977 

Pinus  pondorosa 

'188 

(^larpv/  ^inH  Athorc  1Q7^ 

'30-150  (188) 

Pfister  and  others  1 975 

490-570 

150-250  (150) 

Whittaker  and  Niering  1975 

50-250  (350+) 

Weaver  and  Forcella  1 977 

Festuca  idahioensis 

235 

2.35 

Collins  and  Weaver  1978 

195 

1.95 

Weaver  and  Collins  1977 

147 

1.47 

Daubenmire  1970 

Festuca  scabrella 

152 

1.52 

Willms  and  others  1986 

Bouteioua  gracilis 

103 

1.03 

Weaver  1 983 

53-95 

0.5-1.0 

Hunt  and  others  1 988 

Atriplex  spp. 

28 

West  1983 

'Since  merchantable  production  and  standing  crop  are  emphasized  in  these  studies,  one  might  expect  the  figures  to  be  50  to  66  percent  of  those  reported  in 
studies  of  total  production  (Weaver  and  Forcella  1977). 

Where  production  and  standing  crop  are  expressed  volumetrically,  masses  were  calculated  using  specific  gravities  of  ABLA=0.38,  PIAL=0.40,  PIEL=0.35, 
PSME=0.45,  PIPO=0.43,  PICO=0.38,  THPL=0.33,  TSHE=0.41  (U.S.  FPL  1974). 


(starvation)  may  exclude,  from  an  altitudinal  zone,  plants 
of  lower  zones.  Average  growing  season  temperatures  in 
the  alpine  (and  for  most  zones  below  it)  are  distinctly 
lower  than  temperatures  in  the  vegetation  zone  below 
(fig.  3).  This  correlation  suggests  the  possibility  of — but 
does  not  prove — significant  effects  of  growing  season 
temperature.  In  support  of  this  h3T)othesis,  production- 
temperature  relations  (discussed  below)  suggest  that  the 
vegetation  of  each  altitudinal  zone  is  especially  adapted 
to  temperatures  occurring  in  its  zone.  In  this  regard,  it 
is  satisfying  to  see  that  optimum  temperatures  for  white- 
bark  pine  photosynthesis  (20-25  °C,  Jacobs  and  Weaver 
1990)  are  similar  to  summer  maximum  temperatures 
(10-25  °C,  Weaver  1990). 

Graphs  of  estimated  soil  water  availability  against  veg- 
etation type  (fig.  4)  indicate  little  or  no  stress  in  alpine, 
whitebark,  subalpine  fir,  and  Douglas-fir  zones.  On  this 
basis  one  might  exclude  droiight  as  a  factor  determining 
whitebark's  upslope  limit.  Due  to  two  modifying  factors, 


this  conclusion  seems  premature.  First,  as  one  moves 
upslope  from  forest  to  alpine  (or  moimtain  meadow), 
winds  increase  and,  with  increasing  wind,  drought  in- 
creases due  to  reduced  water  availability — snow  may  be 
blown  off  site  (Daubenmire  1981)  and  water  in  uninsu- 
lated soils  freezes — and  increasing  water  loss — through 
abrasion  of  cuticle  (Hadley  and  Smith  1987)  and  thinning 
of  the  leafs  boundary  layer  (Gates  and  Papian  1971;  Nobel 
1983).  I  see  winter  wind  effects  as  probable  controllers  be- 
cause trees  invading  (or  planted  into)  mountain  meadows 
are  more  often  desiccated  in  winter  than  stunmer.  The 
wind  effect  hypothesis  is  consistent  with  the  fact  that 
groups  of  trees  sometimes  invade  sites  that  individuals 
cannot  invade  alone  (Armand  1992;  Tranquillini  1979).  In 
such  situations  adjacent  trees  are  more  likely  sheltering 
each  other  from  drying  wind  than  from  lethal  low  tempera- 
tures. Second,  as  one  moves  upslope  the  average  soil  be- 
comes progressively  better  drained  (Weaver  1979),  and 
thus  effective  precipitation  is  a  smaller  fraction  of  total 


145 


10      20  30 


ECOSYSTEM  TYPE 


Figure  1 — Midwinter  low  temperatures  in  10  Rocky  Mountain  ecosystems.  Low  temperatures  (°C 
and  °F)  are  represented  by  absolute  winter  minimums  (A,  ABSMN)  and  average  January  minima 
(B,  AVJMN).  Ecosystem  types  range  altitudinally  from  alpine  (2),  down  through  forests  (8-20)  to 
grass  and  shrublands  (26-36);  specifically,  they  are  alpine  (2),  whitebark  pine  (8),  subalpine  fir  (10), 
Douglas-fir  (16),  cedar-hemlock  (18),  ponderosa  pine  (20),  Idaho  fescue  (26),  bluebunch  wheatgrass 
(28),  gramagrass  (30),  and  desert  shrub  (36).  Large  and  small  circles  represent  data  gathered  east 
and  west  of  the  Rockies,  respectively.  Lines  connect  median  values. 


30      40'         '0  10 

ECOSYSTEM  TYPE 


Figure  2— The  coldest  early  spring  (A,  SPR-FRST)  and  late  fall  (8,  FALL-FRST)  frosts  (°C 
and  "F)  in  10  Rocky  Mountain  ecosystems.  Ecosystems,  symbols,  and  lines  are  as  in  figure  1. 


O         10       20       30       40  O        10       20       30  40 

ECOSYSTEM  TYPE 

Figure  3 — Average  growing  season  temperatures  in  1 0  Rocky  Mountain  ecosystems.  The  first  graph 
(A,  TEMP  gs)  gives  the  average  growing  season  temperature  (°C  and  "F).  Q  gs  in  the  second  graph 
(B)  considers  the  exponential  response  of  physiological  processes  to  increasing  temperature  (Q,q). 
Ecosystem  types,  symbols,  and  lines  are  as  in  figure  1 . 


146 


precipitation  at  higher  than  lower  altitudes;  that  is,  the 
Walter  drought  indices  presented  in  figure  4  probably  over- 
estimate water  availability  on  (high-altitude)  thin-soil  sites. 

Downward  Limits 

I  speculate  that  downward  limits  of  whitebark  pine 
might  be  determined  by  heat,  drought,  or  competition. 
The  following  paragraphs  consider  these  hypotheses. 

As  one  moves  downslope  maximum  temperatures 
(fig.  5)  rise.  That  this  correlate  of  tree  disappearance  is 
not  controlling  is  suggested  by  three  facts.  Summer  aver- 
ages in  vegetation  zones  below  whitebark  pine  (10-15  °C, 


fig.  3)  are  well  below  the  whitebark  pine  optimimi 
(20-25  °C,  Jacobs  and  Weaver  1990),  so  downward  migra- 
tion might  actually  improve  production.  July  maxima 
(20-27  °C,  fig.  5)  in  zones  below  are  near  the  whitebark 
pine  optimvuu  temperature  (20-25  °C,  Jacobs  and  Weaver 
1990)  and  thus  should  not  be  damaging.  Long-term 
maxima  (30-42  °C,  fig.  5)  in  zones  below  do  not  eliminate 
net  photosynthesis  (Jacobs  and  Weaver  1990)  and  thus 
are  probably  not  lethal.  In  addition,  if  whitebark  pine 
trees  are  well  watered,  they  grow  well  as  lawn  trees  in  the 
Agropyron  spicatum  zone  (for  example,  Belgrade,  MT) 
where  temperatures  are  far  higher  than  those  fovmd  on 
sites  whitebark  naturally  occupies. 


ECOSYSTEM  TYPE 

Figure  4 — Drought  severity  in  10  Rocky  Mountain  ecosystems.  The  first  graph  (A) 
gives  the  number  of  dry  months  (DRT-MNTHS).  The  second  (B)  indexes  the  an- 
nual water  deficiency  (W-DFCT,  mm).  Both  were  calculated  after  Walter  (1973). 
The  ecosystem  types,  symbols,  and  lines  are  as  in  figure  1 . 


0       10     20     30     40  0       10     20     30  40 

ECOSYSTEM  TYPE 

Figure  5 — High  temperatures  (°C  and  °F)  recorded  in  10  Rocky  Mountain  ecosys- 
tems. The  first  graph  (A)  represents  absolute  summer  highs  (ABSMX).  The  sec- 
ond (B)  gives  average  July  maxima  (AVJMX).  The  ecosystem  types,  symbols,  and 
lines  are  as  in  figure  1 . 


147 


ECOSYSTEM  TYPE 


Figure  6— Biomass  production  in  10  Rocky  Mountain  ecosystems  (see  also  table  2).  The 
first  graph  (A)  presents  harvestable  production  (HPRD,  g/m^/yr).  The  second  (B)  gives  to- 
tal production  (TPRD,  g/m^/yr).  The  ecosystem  types,  symbols,  and  lines  are  as  in 
figure  1 . 


Drought  likely  sets  the  lower  physiological  limit 
(Hutchinson  1958)  for  whitebark  pine's  downward  exten- 
sion. Walter's  (1973)  index  suggests  the  drought  months 
and  water  deficits  are  near  zero  in  high  forests  (whitebark 
pine,  subalpine  fir,  and  Douglas-fir),  greater  in  ponderosa 
pine  forests  and  grasslands,  and  still  greater  in  Atriplex 
shrublands  (fig.  4).  If  (as  noted  above)  wind  decreases 
and  waterholding  capacity  of  soils  increases  downslope, 
water  conditions  for  whitebark  production  improve  as  one 
moves  downslope;  this  is  consistent  with  the  fact  that 
whitebarks  growing  in  the  subalpine  fir  zone  can  be  stately 
timber  producers  (Pfister  and  others  1977;  Weaver  and 
Dale  1977).  Water  deficits  remain  slight  down  through  the 
subalpine  fir  and  possibly  the  Douglas-fir  zones.  The  com- 
plete absence  of  whitebarks  in  the  ponderosa  pine  and 
grassland  zones,  despite  possible  nutcracker  dispersal 
(Lanner  1990;  Tomback  and  others  1990),  supports  my 
doubt  that  the  tree  is  capable  of  growing  in  the  ponderosa 
pine  and  drier  grassland-desert  zones.  Similarly,  in  the 
geographic  dimension,  the  southern  and  eastern  limits  of 
whitebark  pine  seem  to  appear  where  precipitation  in  its 
altitudinal  zone  becomes  too  low  (Arno  and  Weaver  1990). 

The  lower  realized  limit  (Hutchinson  1958)  of  white- 
bark is  set  by  competition  with  subalpine  fir  and  lodge- 
pole  pine  rather  than  drought.  In  our  region,  evidence  for 
this  fact  lies  in  the  tree's  existence  and  good  growth  on 
sites  in  the  subalpine  fir  zone  (but  apparently  not  in  the 
Douglas-fir  zone)  where  subalpine  fir  and  lodgepole  com- 
petition have  been  removed  by  clearcutting  or  fire.  The 
lower  limit  of  whitebark  is,  then,  set  by  those  factors — 
likely  drought  (thin  soils  and  wind)  and  low  summer  tem- 
perature— that  exclude  subalpine  fir  and  lodgepole  pine 
from  higher  sites.  In  the  geographic  dimension,  to  the 
north  and  west  where  rainfall  is  higher,  whitebark  is  also 
unable  to  form  pure  stands  or  disappears  (Arno  and 
Weaver  1990).  (European  foresters  ask  why  we  don't 
eliminate  competitive  "natives"  [subalpine  fir  and  lodgepole 
pine]  from  their  zone  to  allow  expansion  of  whitebark  pine 
and  increased  production  of  quality  white  pine  lumber.) 


Production 

As  one  moves  from  dry  grasslands  to  moist  forests  and 
upward  to  alpine  sites  one  sees  production  rise  and  fall 
again.  Harvestable  production — merchantable  standing 
crop  divided  by  years  in  its  production — rises  from  dry 
grasslands  (100  g/mVyr)  to  moist  forests  (300  g/mVyr)  and 
falls  through  high  forests  to  the  alpine  (100  g/mVyr,  fig.  6, 
table  2).  Available  data  suggest  that  total  aboveground 
production  (total  increment  ^ro?lyr)  also  increases  fi-om 
dry  grass  and  shrublands  (50-100  g/mVjr)  to  warm  forests 
(over  500  g/mVyr)  and  falls  to  the  alpine  (100  g/mV5T, 
fig.  6,  table  1).  The  trend  is  parallel,  but — at  least  in  for- 
ests— total  production  is  far  higher  than  harvestable  pro- 
duction. Three  factors  may  contribute  to  this  difference: 
(1)  Much  of  the  production,  over  the  life  of  a  stand,  is  lost 
to  needle  drop,  self  pruning,  and  natural  thinning;  this 
material  decomposes  and  is  recycled.  The  total  produc- 
tion rates  reported  are  deceptively  high  because  they  in- 
clude nutrients  being  recycled  to  the  atmosphere  and  soil; 
that  is,  this  production  consists  of  an  actual  production 
component  based  on  nutrient  import  (weathering  and  at- 
mospheric delivery)  and  a  restructuring  component  based 
on  recycling  of  nutrients  from  "obsolete"  leaves,  branches, 
and  trees.  (2)  Only  33  to  50  percent  of  the  material  extant 
at  harvest  is  merchantable  (Weaver  and  Forcella  1977). 
(3)  Most  of  the  total  production  data  comes  from  areas 
outside  our  region.  If  the  climates  of  Douglas-fir,  cedar- 
hemlock,  and  ponderosa  pine  stands  from  which  produc- 
tion data  are  reported  are  moister  or  warmer  than  the 
Montana  stands  in  which  the  harvestable  productivity 
data  were  gathered,  somewhat  higher  productivities 
might  be  expected. 

I  expect  productivity  to  be  controlled,  like  vegetation 
distribution,  by  climate.  In  the  following  paragraphs  I 
will  explore  the  h3rpothesis  that  production  depends  on 
growing  season  length  and  the  warmth  of  that  season. 

Growing  season  length  might  be  indexed  as  the  number 
of  months  where  daily  average  temperatures  are  above 


148 


Table  3 —  Relationship  of  production  (g/m^/yr)  to  climatic  factors 


Harvestable  production' 


HP 

66.00 

+ 

12.8000 

WS 

p  =  0.50 

=  0.05 

-171.00 

+ 

70.0000 

GS 

p  =  0.00 

=  0.86 

34.00 

+ 

1.7700 

GSxT 

p  =  0.04 

/■2  =  0.44 

-38.00 

+ 

14.4000 

GSxQ 

p  =  0.01 

=  0.58 

Log  HP  = 

0.78 

wo 

n  -  n  nn 

f2  _  n  7fi 

1  75 

n  nnfin 

n  —  n  1 5 

f2  —  n  97 

1.44 

+ 

0.0530 

GSxQ 

p  =  0.03 

r2  =  0.44 

tal  Droduction^ 

TP 

-880.00 

+ 

321.0000 

GS 

p  =  0.16 

r2  =  0.22 

50.00 

+ 

8.3200 

GSxT 

p  =  0.32 

=  0.12 

-398.00 

+ 

77.0000 

GSxQ 

p  =  0.19 

r2  =  0.20 

Log  TP  = 

0.40 

+ 

0.4700 

GS 

p  =  0.03 

=  0.45 

1.96 

+ 

0.0086 

GSxT 

p  =  0.32 

r2  =  0.12 

1.47 

+ 

0.0820 

GSxQ 

p  =  0.22 

^  =  0.22 

'Harvestable  production  (HP,  lop  diameter  greater  than  10  cm)  and  the  log  of  harvestable  production 
(log  HP)  are  regressed  against  warm  season  (WS),  warm-moist  growing  season  (GS),  the  product  of 
warm-moist  growing  season  and  average  growing  season  temperature  (GSxT),  and  the  product  of 
wann-moist  growing  season  and  Q,o-based  growth  support  units  (GSxQ). 

n"otal  production  (TP)  and  log  total  production  (log  TP)  are  also  regressed  against  GS,  GSxT,  and 
GSxQ. 


0  °C.  The  rationale  is  that,  iinless  air  temperatiires  are 
above  0  °C,  exposed  soils  will  be  frozen  (Weaver,  in  prepa- 
ration) and  frozen  soils  will  deliver  little  water  and  nutri- 
ents. Because  a  regression  of  production  against  this  in- 
dex of  growing  season  is  so  loose  (r^  =  0.05,  table  3,  fig.  7), 
the  hypothesis  that  warm-season  length  controls  produc- 
tion is  inadequate. 

The  previous  index  may  fail  because,  at  least  at  lower 
altitudes,  growing  season  is  limited  in  late  summer  by  a 
lack  of  soil  water,  rather  than  temperatiire.  If  so,  the 
growing  season  index  would  be  improved  as  an  index  of 
production  by  subtracting  soil-dry  months  from  total  soil- 
warm  months.  As  above,  I  index  drought  as  periods  when 
precipitation  (mmy2)  is  less  than  average  temperature  (C, 
after  Walter  1973).  Regressions  of  production  against 
months  without  either  temperature  or  moisture  stress  ex- 
plain most  of  the  variance  in  harvestable  production  (r^  = 
0.86,  table  3,  fig.  7).  Season  length  is  a  weaker  correlate 
of  total  production  (r^  =  0.22).  I  offer  two  unsatifying 
hypotheses  for  this  important  difference:  (1)  Growing  sea- 
son predicts  harvestable  production  better  than  total  pro- 
duction because  photosynthates  are  allocated  first  to 
canopy  maintainence  (so  this  component  does  not  vary 
much  among  types)  and  second  to  supporting  structure;  if 
so,  the  relationship  between  season  length  and  tnmk  pro- 
duction is  clarified  by  omitting  the  common  leaf-twig  pro- 
duction. (2)  If  total  production  were  measured  in  slightly 
more  favorable  segments  of  the  environmental  types  than 
harvestable  production,  the  iinderstatement  of  season 
length  would  weaken  the  relationship. 

If  all  vegetation — grassland  to  forest  to  alpine — had  a 
single  temperature  response  curve,  we  would  expect  pro- 
duction in  "warm  moist  soil  days"  to  rise  exponentially 
with  increasing  temperatiu-e.  This  expectation  is  based 


on  the  rise  in  chemical  reaction  rates  with  rising  tem- 
perature, on  the  resultant  physiological  response  curves 
(Larcher  1975),  on  degree-day  prediction  of  plant  phenol- 
ogy (Chang  1960),  and  on  the  prediction  that  degree-day 
formulae  might  be  improved  by  use  of  temperatiire  aver- 
ages computed  from  a  nonlinear  Q^^  relationship  (Larcher 
1975).  Thus,  we  expect  production  to  be  better  correlated 
with  the  product  of  growing  season  days  x  temperature 
than  with  growing  season  days  alone.  In  fact,  whether 
we  use  our  average  growing  season  temperature  or  grow- 
ing season  Q  index,  the  relationship  is  poorer  for  both 
harvestable  (r^=  0.44)  and  total  (r^=  0.12)  production. 
Since  the  basic  physiological  response  is  a  physiochemical 
necessity,  I  deduce  first  that,  while  the  vegetation  of  each 
environmental  zone  does  respond  to  temperature,  it  is 
specifically  adapted  to  temperattires  in  that  zone  and,  sec- 
ond, that  temperature  adaptation  distinct  to  each  zone 
eliminates  temperature  fi'om  the  predictive  equation.  I 
doubt  that  temperature  would  be  eliminated  if  adaptation 
were  eliminated  by  using  a  genetically  homogeneous  veg- 
etation type;  for  example,  if  the  region  were  vegetated 
with  a  single  crop,  production  would  be  best  predicted  by 
the  product  of  growing  season  length  and  temperature. 
Since  competitiveness  of  a  tree  undoubtedly  depends  on 
its  productivity  relative  to  competitors,  this  conclusion 
supports  earlier  speculation  that  growing  season  tempera- 
ture adaptation  provides  at  least  one  basis  for  the 
existance  of  different  vegetation  zones. 

While  I  have  argued  that  production  in  a  zone  is  best 
predicted  from  the  length  of  its  warm  moist  growing  sea- 
son, I  expect  the  standing  crop  of  mature  accvmaulative 
(woody)  vegetation  (table  2)  to  be  largely  determined  by 
site  fertility.  Accmnulation — slow  or  fast — depends  on 
photosynthesis  minus  respiration.  Synthesis  depends  on 


149 


light,  warmth,  water  (open  stomates),  and  nutrients.  In 
the  absence  of  stand  destruction,  annual  inputs  of  light, 
warmth,  water,  and  elements  with  atmospheric  cycles 
(for  example,  C,  H,  O,  N)  should  support  eternal  accumu- 
lation. Limited  supplies  of  nutrients  with  geologic  cycles 
(for  example,  P,  K,  Ca)  will  halt  accumulation  (Weaver 
1978).  It  is  also  argued  (Odum  1969)  that  accvunulation 
is  halted  when  respiration  equals  production.  Thus,  in 
young  trees  net  production  is  high  because  the  ratio  of 
photosynthetic  mass  to  respiratory  mass  is  high,  in  longer 
stemmed  trees  efflcency  declines  as  the  respiratory  load 
increases,  and  growth  ceases  when  respiration  equals 
gross  production;  heartwood's  contribution  to  this  rela- 
tionship is  in  proportion  to  its  inertness.  While  the  respi- 
ratory factor  undoubtedly  contributes  to  observed  declines 
in  production  with  stand  height  (and  age)  (Weaver  and 
others  1990),  I  see  nutrient  supplies  as  more  limiting  in 
high  forests  because  the  trees  are  far  shorter  (and  there- 
fore more  efficient)  than  productive  trees  of  lower  alti- 
tudes. In  contrast,  standing  crops  in  less  accumulative 
(herbaceous)  vegetation  are  determined  by  production  oc- 
curring in  one  growing  season,  and  while  it  could  be  de- 
termined by  supplies  of  a  geologically  cycling  nutrient, 
standing  crop  is  more  likely  determined  by  a  bulk  re- 
source like  light  (unlikely),  warmth,  water,  or  an  atmo- 
spheric nutrient  like  nitrogen. 


CONCLUSIONS 

Whitebark  pine  vegetation  is  likely  excluded  from  the  al- 
pine zone  by  cool  growing  season  temperatiires  or  droughts 
occurring  most  likely  in  winter.  It  seems  unlikely  that 
freezes  of  late  fall,  midwinter,  or  early  spring  exclude  the 
tree  from  higher  sites. 

The  lower  physiological  limit  of  whitebark  pine  is  likely 
set  by  drought.  Its  lower  realized  limit  is  likely  due  to  com- 
petition with  lodgepole  pine  or  subalpine  fir.  By  control  of 
competition,  managers  could  probably  extend  whitebark's 
range  downslope. 

The  distribution  of  other  dominant  plants  on  the  tem- 
perature water  gradient  may  be  similarly  controlled. 

Annual  production  is  strongly  related  (H  =  0.86)  to  the 
number  of  growing  season  days.  The  lack  of  correlation 
with  temperature  suggests  that  plants  of  any  vegetation 
zone  are  adapted  to  temperature  conditions  peculiar  to 
that  zone.  Mature  standing  crops  of  woody  vegetation  are 
more  likely  determined  by  nonatmospheric  nutrients  than 
temperature  and  precipitation.  Potential  standing  crops 
of  herbaceous  vegetation,  on  the  other  hand,  are  more 
likely  limited  by  bulk  resources  such  as  warmth,  water, 
carbon,  or  nitrogen. 


3  6 

WM  SEASON,  mo 


20  40 

GS  X  T  (moxC  5) 


\ 

o 

cr 

CL 
X 


r2=  0.85 


B 


J  L 


J  L 


J  L 


3  6 

GRO  SEASON,  mo 


GS  X  Q  (moxQ) 


Figure  7 — The  relationship  of  harvestable  production  (HPROD,  g/m^/yr)  to  season 
length  and  temperature:  (A)  production  vs.  warm  season  (WM-SEASON),  (B)  produc- 
tion vs.  warm-moist  season  (GRO-SEASON),  (C)  production  vs.  warm-moist  season 
(GS)  X  average  summer  temperature  above  5  °C  (C-5),  (D)  production  vs.  warm  moist 
season  (GS)  x  growth  support  units  (Qg^).  The  ecosystem  types  are  as  in  figure  1 . 


150 


ACKNOWLEDGMENTS 

I  thank  colleagues  who  helped  me  locate  weather  sta- 
tions representative  of  major  environmental  types  of  the  re- 
gion (S.  Amo,  D.  Despain,  J.  Habeck,  R.  Rankin,  and  others 
acknowledged  in  Weaver  1980  and  1990),  acquire  the  data 
(J.  Caprio),  and  choose  parameters  for  study  (J.  Pickett  and 
J.  Brown).  The  critical  review  of  P.  Burgeron,  D.  Despain, 
S.  Harvey,  W.  Moir,  and  D.  Perry  is  greatly  appreciated.  I 
am  also  grateful  for  the  support  provided  by  Intermoimtain 
Research  Station,  USDA  Forest  Service  (INT-92720- 
RJVA). 

REFERENCES 

Aplet,  G.;  Smith,  F.;  Laven,  R.  1989.  Stemwood  biomass 
and  production  during  spruce-fir  stand  development. 
Journal  of  Ecology.  77:  70-77. 

Armand,  A.  1992.  Sharp  and  gradual  motmtain  timber- 
lines  as  a  result  of  species  interaction.  In:  Hansen,  A.; 
DiCastri,  F.  1992.  Landscape  boundaries.  New  York: 
Springer  Verlag:  360-378. 

Amo,  S.  1984.  Timber  line — moimtain  and  arctic  forest 
frontiers.  Seattle:  Moimtaineers.  304  p. 

Amo,  S.;  Weaver,  T.  1990.  Whitebark  pine  commimity 
types  and  their  pattem  in  the  landscape.  In:  Schmidt, 
W.;  McDonald,  K,  comps.  Proceedings —  symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-moimtain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermovmtain  Research  Station:  97-105. 

Burke,  M.;  Gusta,  L.;  Quamme,  H.;  Weiser,  C.;  Li,  P. 
1976.  Freezing  injxiry  in  plants.  Annual  Review  of  Plant 
Physiology.  27:  507-528. 

Chang,  J.  1968.  Climate  and  agriculture.  Chicago:  Aldine. 
304  p. 

Collins,  D.;  Weaver,  T.  1978.  Effects  of  summer  weather 
modification  (irrigation  in  Festuca  idahoensis-Agro- 
pyron  spicatum  grasslands.  Jovu-nal  of  Range  Manage- 
ment. 31:  264-269. 

Daubenmire,  R.  1943.  Vegetation  zonation  in  the  Rocky 
Mountains.  Botanical  Review.  9:  325-393. 

Daubenmire,  R.  1956.  Climate  as  a  determinant  of  vegeta- 
tion distibution  in  eastern  Washington  and  northern 
Idaho.  Ecological  Monographs.  26:  131-154. 

Daubenmire,  R.  1970.  Steppe  vegetation  of  Washington. 
Bull.  62.  Pullman,  WA:  Washington  Agricultural  Ex- 
periment Station.  131  p. 

Daubenmire,  R.  1981.  Subalpine  parks  associated  with 
snow  transfer  in  the  mountains  of  northern  Idaho  and 
eastem  Washington.  Northwest  Science.  55:  124-135. 

Daubenmire,  R.;  Daubenmire,  J.  1968.  Forest  vegetation 
of  eastern  Washington  and  northern  Idaho.  Bull.  60. 
Pullman,  WA:  Washington  Agricultural  Experiment 
Station.  104  p. 

Forcella,  F.;  Weaver,  T.  1977.  Biomass  and  productivity  of 
the  subalpine  Pinus  albicaulis/Vaccinium  scoparium  as- 
sociation, MT  USA.  Vegetatio.  35:  95-105. 

Gams,  H.  1931.  Die  klimatische  Begrenzimg  von  Pflanzen 
arealen  und  die  Verteilvmg  der  hygrischen  Konti- 
nentalitat  in  den  Alpen.  Zietschrit  der  Gesellschafl  fur 
Erdkunde.  9:  32-68. 


Gates,  D.;  Papian,  L.  1971.  Atlas  of  energy  budgets  of 
plant  leaves.  London  and  New  York:  Academic  Press. 
277  p. 

Geiger,  R.  1965.  The  climate  near  the  ground.  Cambridge, 

MA:  Harvard  University  Press.  611  p. 
Hadley,  J.;  Smith,  W.  1987.  Influence  of  knunmholz 

microclimate  on  needle  physiology  and  survival. 

Oecologia.  73:  82-90. 
Hanley,  D.  1976.  Tree  biomass  and  production  estimates 

for  three  habitat  types  in  northern  Idaho.  Bull.  14. 

Moscow,  ID:  University  of  Idaho,  College  of  Forestry. 

16  p. 

Hunt,  W.;  Ingham,  E.;  Coleman,  D.;  ElHott,  E.;  Reid,  P. 
1988.  Nitrogen  limitation  of  production  and  decomposi- 
tion in  prairie,  moimtain  meadow,  and  pine  forest. 
Ecology.  69:  1009-1016. 

Holdridge,  L.  1967.  Determination  of  world  plant  forma- 
tions from  simple  climatic  data.  Science.  130:  572. 

Huschle,  G.;  Hironaka,  M.  1980.  Classification  and  ordi- 
nation of  serai  plant  communities.  Journal  of  Range 
Management.  33:  179-182. 

Hutchinson,  G.  1958.  Concluding  remarks.  Cold  Springs 
Harbor  Symposium  in  Quantiative  Biology.  22:  415-427. 

Jacobs,  J.;  Weaver,  T.  1990.  Effect  of  temperature  and 
temperature  preconditioning  on  seedlimg  performance 
of  whitebark  pine.  In:  Schmidt,  W.;  McDonald,  K, 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S. 
Department  of  Agriculture,  Forest  Service,  Intermoun- 
tain  Research  Station:  134-139. 

Kuchler,  A.  1964.  Potential  vegetation  of  the  contermi- 
nous US.  Spec.  Publ.  36.  New  York:  American  Geo- 
graphical Society.  155  p. 

Landis,  T.;  Mogren,  E.  1975.  Tree  strata  biomass  of  sub- 
alpine spmce-fir  stands  in  SW  Colorado.  Forest  Science. 
21:  9-12. 

Lanner,  R.  1980.  Biology,  taxonomy,  evolution,  and  geog- 
raphy of  the  stone  pines  of  the  world.  In:  Schmidt,  W.; 
McDonald,  K.,  comps.  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  14-24. 

Larcher,  W.  1975.  Physiological  plant  ecology.  New  York: 
Springer.  252  p. 

Neilson,  R.  1986.  High  resolution  climatic  analysis  and 
southwest  biogeography.  Science.  232:  27-33. 

Nielson,  R.  1992.  Toward  a  rule  based  biome  model.  Land- 
scape Ecology.  7:  27-43. 

Nobel,  P.  1983.  Biophysical  plant  physiology.  San 
Francisco:  Freeman.  608  p. 

Penman,  H.  1949.  The  dependence  of  transpiration  on 
weather  and  soil  conditions.  Journal  of  Soil  Science. 
1:  74-89. 

Pfister,  R.;  Kovalchik,  B.;  Amo,  S.;  Presby,  R.  1977.  For- 
est habitat  types  of  Montana.  Gen.  Tech.  Rep.  INT-34. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Forest  and  Range  Experiment  Sta- 
tion. 174  p. 


151 


Rodin,  L.;  Bazilevich,  N.  1967.  Production  and  mineral  cy- 
cling in  terrestrial  vegetation.  London:  Oliver  and  Boyd. 
288  p. 

Salisbury,  F.;  Ross,  C.  1992.  Plant  physiology.  Belmont, 
CA:  Wadsworth.  682  p. 

Scott,  D.;  Billings,  W.  1964.  Effects  of  environmental  fac- 
tors on  standing  crop  and  production  of  an  alpine  tun- 
dra. Ecological  Monographs.  34:  243-270. 

Stephenson,  N.  1990.  Climatic  control  of  vegetation  distri- 
bution: the  role  of  water  balance.  American  Naturalist. 
135:  649-670. 

Thilenius,  J.;  Smith,  D.;  Brown,  G.  1974.  Effect  of  2,4-D 
on  composition  and  production  of  an  alpine  plant  com- 
munity in  Wyoming.  Journal  of  Range  Management. 
27:  140-142. 

Thornthwaite,  C.  1948.  An  approach  toward  a  rational 
classification  of  climate.  Geographical  Review.  38: 
55-94. 

Tomback,  D.;  Hoffmann,  L.;  Sund,  S.  1990.  Coevolution  of 
whitebark  pine  and  nutcrackers:  implications  for  forest 
regeneration.  In:  Schmidt,  W.;  McDonald,  K.,  comps. 
Proceedings — sjrmposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source. Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  De- 
partment of  Agriculture,  Forest  Service,  Intermountain 
Research  Station:  118-129. 

Tranquillini,  W.  1979.  Physiological  ecology  of  the  alpine 
timberline.  Tree  existence  at  high  altitude  with  special 
reference  to  the  European  Alps.  New  York:  Springer. 
137  p. 

U.S.  FPL.  1974.  Wood  handbook:  wood  as  an  engineering 
material.  Agric.  Handb.  72.  Washington,  DC:  U.S.  De- 
partment of  Agriculture,  Forest  Service,  Forest  Prod- 
ucts Laboratory,  [n.p.]. 

USGS.  1970.  The  national  atlas.  Washington,  DC:  U.S. 
Department  of  the  Interior,  Geological  Survey.  417  p. 

Walter,  H.  1973.  Vegetation  of  the  earth  in  relation  to  cli- 
mate and  ecophysiological  conditions.  New  York: 
Springer.  237  p. 

Weaver,  T.  1978.  Changes  in  soils  along  a  vegetation- 
altitudinal  gradient  of  the  Northern  Rocky  Mountains. 
In:  Youngberg,  C,  ed.  Forest  soils  and  land  use.  Pro- 
ceedings; 5th  North  American  Soils  Conference.  Fort 
Collins,  CO:  Colorado  State  University,  Forestry 
Department:  14-29. 


Weaver,  T.  1980.  Climates  of  vegetation  types  of  the 
northern  Rocky  Mountains  and  adjacent  plains.  Ameri- 
can Midland  Naturalist.  103:  392-398. 

Weaver,  T.  1983.  Yield  response  of  high  plains  grasslands 
to  water  enrichment,  three  phases.  In:  Holman,  L.; 
Knutsen,  G.  State  of  Montana  activities  in  the  High 
Plains  Cooperative  Program  1981-1983.  Helena,  MT: 
Montana  Department  of  Natural  Resources,  Water 
Resources  Division:  77-94. 

Weaver,  T.  1990.  Climates  of  subalpine  pine  woodlands. 
In:  Schmidt,  W.;  McDonald,  K.,  comps.  Proceedings — 
S3mtiposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
72-79. 

Weaver,  T.;  Collins,  D.  1977.  Effects  of  weather  modifica- 
tion (increased  snow  pack)  on  Festuca  idahoensis  mead- 
ows. Journal  of  Range  Management.  30:  451-456. 

Weaver,  T.;  Dale,  D.  1974.  Pinus  albicaulis  in  central 
Montana:  environment,  vegetation,  and  production. 
American  Midland  Naturalist.  92:  222-229. 

Weaver,  T.;  Forcella,  F.  1977.  Biomass  of  fifty  conifer  for- 
ests and  nutrient  exports  associated  with  their  harvest. 
Great  Basin  Naturahst.  37:  395-401. 

Weaver,  T.;  Forcella,  F.;  Dale,  D.  1990.  Stand  develop- 
ment in  whitebark  pine  woodlands.  In:  Schmidt,  W.; 
McDonald,  K.,  comps.  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  151-155. 

West,  N.  1983.  Temperate  deserts  and  semideserts:  Inter- 
mountain salt  desert  shrubland.  Amsterdam:  Elsevier: 
382. 

Whittaker,  R.  1975.  Communities  and  ecosystems.  New 

York:  Macmillan.  385  p. 
Whittaker,  R.;  Niering,  W.  1975.  Vegetation  of  the  Santa 

Catalina  Mountains,  Arizona.  V.  Biomass,  production 

and  diversity  along  the  elevation  gradient.  Ecology.  56: 

771-790. 

Willms,  W.;  Smoliak,  S.;  Bailey,  A.  1986.  Herbage  produc- 
tion following  litter  removal  on  Alberta  native  grass- 
lands. Journal  of  Range  Management.  39:  536-540. 


152 


Regeneration 


153 


A 


VARIATION  IN  SIZE  AND  WEIGHT  OF 
CONES  AND  SEEDS  IN  FOUR  NATURAL 
POPULATIONS  OF  CARPATHIAN 
STONE  PINE 

I.  Blada 
N.  Popescu 


Abstract — High  variation  was  found  within  populations  of 
stone  pine  (Pinus  cembra  L.)  for  seeds  per  cone  and  seeds  per 
cone  weight,  but  middle  high  for  1,000-seed  weight  and  low  for 
cone  length  and  cone  diameter.  The  cone  length,  cone  diameter, 
seeds  per  cone,  seeds  per  cone  weight,  and  1,000-seed  weight 
means  were  4.7  cm,  4.2  cm,  37.8  cm,  9.2  g,  and  250  g,  respec- 
tively. Southern  populations  were  similar  in  cone  length  and 
cone  diameter,  but  not  in  seeds  per  cone,  seeds  per  cone  weight, 
and  l.OOO-seed  weight.  Cone  length  and  cone  diameter  from 
open-pollinated,  cross-pollinated  and  self-pollinated  cones  were 
similar,  while  seeds  per  cone,  seeds  per  cone  weight,  and  1,000- 
seed  weight,  were  not.  All  traits  displayed  continuous  variation. 
Significant  correlations  were  found  among  all  traits,  except 
1,000-seed  weight. 


Stone  pine  (Pinus  cembra  L.)  is  naturally  distributed 
in  the  highest  forest  zone  of  the  Alps  and  Carpathian 
Mountains  (Critchfield  and  Little  1966;  Holzer  1975). 
In  the  Alps,  the  low-elevation  stands  range  between  1,100 
to  1,500  m,  but  the  main  zone  extends  between  1,700  to 
2,000  m  (Contini  and  Lavarelo  1982;  Holzer  1975),  while 
the  high-elevation  form  of  the  species  climbs  as  single 
trees  up  to  2,700  m  above  sea  level  (Moser  1960).  In 
Romania,  stone  pine  ranges  from  1,350  to  1,880  m  in  the 
northern  Carpathians  (Gubesch  1971)  and  from  1,350  to 
1,980  m  in  the  southern  Carpathians  (Beldie  1941;  Oarcea 
1966;  Tataranu  and  Costea  1952). 

Because  of  its  tolerance  for  low  temperatvires,  the  spe- 
cies is  very  important  for  reforestation  of  the  subalpine 
zone;  in  this  zone  it  is  also  important  on  watersheds,  for 
stabilizing  avalanche  areas,  and  for  reducing  the  effects 
of  flash  floods  (Holzer  1972). 

Stone  pine  has  a  particular  importance  for  the  silvicul- 
ture of  the  subalpine  zone  of  the  Carpathians.  For  this 
reason  a  genetic  improvement  program  with  both  intra- 
and  interspecific  crosses  is  being  carried  out  (Blada  1982, 
1990a);  some  results  have  been  published  (Blada  1987, 
1990b,  1992a)  or  are  in  preparation  (Blada  1992b). 

This  paper  reports  on  the  phenotypic  variation  in  size 
and  weight  of  cones  and  seeds  in  four  natural  populations 
of  Carpathian  stone  pine,  as  part  of  the  program. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

I.  Blada  and  N.  Popescu  are  Forest  Geneticists,  Forest  Research  Insti- 
tute, Bucharest  11,  Romania. 


MATERIALS  AND  METHODS 

Twenty  trees  were  sampled  at  random  in  each  of  the 
four  populations  listed  in  table  1.  In  mid-July  1991,  25 
cones  from  open  pollination  (OP)  on  each  of  the  20  trees 
(total  80  trees)  were  protected  against  the  mountain  jay 
(Nucifraga  caryocatactes  L.)  by  using  metal  net  bags  25 
by  20  cm  in  size. 

For  comparative  reasons,  cones  and  seeds  obtained  from 
controlled  cross-  and  self-pollinations  (CP,  SP)  in  the 
Gemenele  population  were  measured;  these  pollinations 
were  performed  in  July  1989  using  10  x  10  full  diallel 
mating  design  (Blada  1992b),  according  to  Griffing's 
(1956)  Experimental  Method  1. 

Cone  and  seed  measurements  were  taken  soon  afler 
their  collection  in  October  1990  for  controlled  cross- 
pollinated  and  self-pollinated  and  in  October  1991  for 
open-pollinated  cones.  Twenty-five  cones  were  measured 
from  each  tree  resulting  fi*om  open-pollinated  cones,  and 
20  cones  fi-om  each  combination  resulting  fi*om  controlled 
cross-pollinated  and  self-pollinated  cones.  Five  traits 
were  measured  as  shown  in  table  2. 

Using  data  from  measurements,  the  population  mean 
(x),  standard  deviation  (g),  mean  square  {&),  variation  co- 
efficient (VC),  range  of  variation  (Q),  and  correlation  coef- 
ficient (r)  were  calculated. 

WITHES-POPULATION  VARIATION 

Statistical  parameters  of  the  cones  and  seeds  of  the  four 
Carpathian  stone  pine  populations  listed  in  table  1  are 
summarized  in  table  3  and  figure  1. 


Table  1 — Geographic  distribution  of  Pinus  cembra  studied 
populations  in  the  Carpathians 


Population 

Latitude  N.  Longitude  E. 

Altitude 

Carpathian 
Range 

 Degrees  

Meters 

Gemenele 

45°35"  22°50' 

1,720 

Southern 

Stana  de  Rau 

45°25'  23°03' 

1,450 

Southern 

Pietrele 

45°23'  22°52' 

1,550 

Southern 

Lala 

47°33'  25°05' 

1,520 

Northern 

154 


Table  2 — Measured  traits 


Trait  Units  Symbols 


Cone  length  Centimeters  CL 

Cone  diameter  Centimeters  CD 

Seeds  per  cone  Number  SC 

Seeds  per  cone  weight  Grams  SCW 

1 ,000-seed  weight  Grams  1 ,000-SW 


The  following  main  results  were  obtained  for  the 
Gemenele,  Stana  de  Rau,  Pietrele,  and  Lala  populations 
(in  sequence): 

•  The  cone  length  mean  was  4.9  ±  0.4  cm,  4.7  ±  0.4  cm, 
4.9  ±  0.6  cm,  and  4.1  ±  0.4  cm,  respectively. 

•  The  cone  diameter  mean  was  4.2  ±  0.3  cm,  4.1  ±  0.3  cm, 
4.1  ±  0.3  cm,  and  3.5  ±  0.2  cm,  respectively. 

•  The  seeds  per  cone  mean  was  40.9  ±  9.6,  36.7  ±  8.1, 
52.6  ±  13.3,  and  21.1  ±  7.1,  respectively. 

•  The  seeds  per  cone  weight  was  9.8  ±  2.7  g,  9.0  ±  2.5  g, 
12.5  ±  3.9  g,  and  5.6  ±  1.7  g,  respectively. 


Table  3 — Mean  (x),  phenotypic  standard  deviation  (a),  mean  square 
(cf),  variation  coefficient  {VQ,  range  of  variation  (O) 


Parameters  (open  pollination) 


Trait 

x±  a 

VC 

Q 

Gemenele  population 

CL 

4.9 

±  0.4 

0.15 

7.9 

4.1 

5.8 

CD 

4.2 

+  0.3 

0.11 

7.9 

3.6 

5.0 

SC 

40.9 

±  9.6 

91.31 

23.3 

22.5 

60.2 

SCW 

9.8 

±  2.7 

7.50 

28.0 

4.6 

18.1 

1,000-SW 

238.0 

±  47.4 

2,243 

19.9 

170 

352 

Stana  de  Rau  population 

CL 

4.7 

±  0.4 

0.18 

9.1 

3.8 

5.6 

CD 

4.1 

±  0.3 

0.08 

7.0 

3.7 

4.9 

SC 

36.7 

±  8.1 

65.67 

22.1 

21.7 

55.4 

SCW 

9.0 

±  2.5 

6.05 

27.3 

5.5 

14.1 

1,000-SW 

252 

±  37.6 

1,412 

14.9 

175 

293 

Pietrele  population 

CL 

4.9 

±  0.6 

0.34 

11.7 

4.1 

6.0 

CD 

4.1 

±  0.3 

0.10 

7.7 

3.4 

4.7 

SC 

52.6 

±  13.3 

177.31 

25.3 

31.2 

90.4 

SCW 

12.5 

±  3.9 

15.27 

31.3 

6.2 

23.1 

1 ,000-SW 

238 

±  60.9 

3,712 

25.6 

138 

428 

Lala  population 

CL 

4.1 

±  0.4 

0.17 

9.9 

3.7 

5.1 

CD 

3.5 

±  0.2 

0.04 

6.0 

3.1 

3.9 

SC 

21.1 

±  7.1 

51.15 

34.0 

13.6 

34.8 

SCW 

5.6 

±  1.7 

2.78 

29.9 

3.6 

9.4 

1 ,000-SW 

270 

±  46.5 

2,163 

17.2 

184 

346 

Trait  parameters  across  four  populations 

CL 

4.7 

±  0.5 

0.30 

11.7 

3.7 

6.0 

CD 

4.2 

±  0.3 

0.11 

7.9 

3.1 

5.0 

SC 

37.8 

±  14.9 

222.22 

39.4 

13.6 

90.4 

SCW 

9.2 

±  3.7 

13.77 

40.3 

3.6 

23.1 

1 ,000-SW 

250 

±  49.7 

2,471.00 

19.8 

167.0 

354.0 

•  The  1,000-seed  weight  mean  was  238  +  47.4  g,  252  ± 
37.6  g,  238  ±  60.9  g,  and  270  ±  46.5  g,  respectively. 

Therefore,  the  southern  populations  were  similar  in 
cone  length  and  cone  diameter,  but  not  in  seeds  per  cone, 
seeds  per  cone  weight,  and  1,000-seed  weight.  On  the 
other  hand,  the  southern  populations  ranked  high  and  the 
northern  population  ranked  low  for  all  measured  traits. 

Very  high  variation  coefficients  were  found  within 
each  population  for  both  seeds  per  cone  and  seeds  per 
cone  weight,  but  middle  high  for  1,000-seed  weight  and 
low  for  cone  length  and  cone  diameter  traits. 

The  last  three  columns  of  table  3  give  information  con- 
cerning mean  squares,  variation  coefficients,  and  the 
range  of  variation  for  pop\ilation  traits. 

ACROSS-POPULATION 
PARAMETERS 

The  mean  values  and  ranges  of  the  traits  across  foiu* 
studied  populations — and  by  extrapolation — for  all  the 
Carpathian  stone  pine  were  as  follows  (table  3,  lower 
part): 

•  The  cone  length  mean  ranged  from  3.7  to  6.0  cm,  with 
a  mean  of  4.7  ±  0.5  cm,  and  the  cone  diameter  ranged 
from  3.1  to  5.0  cm,  with  a  mean  of  4.2  ±  0.3  cm.  According 
to  Contini  and  Lavarelo  (1992),  the  size  of  cones  from  the 
Alps  ranged  from  5.0  to  10.0  cm  in  length  and  from  4.0  to 
6.0  cm  in  diameter.  Therefore,  the  cone  size  from  the  Alps 
ranked  high  and  the  cones  from  the  Carpathians  ranked 
low. 

•  The  seeds  per  cone  ranged  from  13.6  to  90.4,  with  a 
mean  of  37.8  ±  14.9.  However,  the  seeds  per  cone  mean 
from  the  Alps  varied  between  46  and  164  with  a  mean 
of  93  seeds  (Rohmeder  and  Rohmeder  1955).  Thus,  the 
seeds  per  cone  from  the  Alps  ranked  high  and  the  seeds 
per  cone  from  the  Carpathians  ranked  low. 

•  The  seeds  per  cone  weight  mean  ranged  from  3.6  to 
23.1  g,  with  a  mean  of  9.2  ±  3.7  g. 

•  The  1,000-seed  weight  mean  ranged  from  167  to 
354  g,  with  a  mean  of  250  ±  49.7  g.  Consequently,  a  kilo- 
gram of  seed  from  the  Carpathians  could  include  from 
2,825  to  5,988  seeds,  with  a  mean  of  4,000  seeds.  Accord- 
ing to  Rohmeder  and  Rohmeder  (1955),  the  1,000-seed 
weight  mean  from  the  Alps  ranged  between  150  and 
350  g;  thus,  the  Carpathians  1,000-seed  weight  mean 
was  very  close  to  that  of  the  Alps. 

The  cone  length,  cone  diameter,  seeds  per  cone,  seeds 
per  cone  weight,  and  1,000-seed  weight  coefficients  ac- 
counted for  11.7  percent,  7.9  percent,  39.4  percent,  40.3 
percent,  and  19.8  percent  of  the  variation,  respectively 
(table  3,  lower  part). 

POLLINATION  COMPARISONS 

The  cone  and  seed  parameters  shown  in  table  4  came 
from  records  of  2,000  cones  from  80  open-pollinated  (OP) 
trees,  1,800  cones  from  a  10  x  10  full  diallel  mating  design 
for  cross-pollination  (CP),  and  200  cones  from  10  self- 
pollinated  (SP)  trees,  all  from  the  Gemenele  population. 
For  example,  the  cone  length  ranged  from: 


155 


3- 


2. 


Oi 

c 


300, 


250. 


200. 


.C150. 

Jioo 


o 
o 

°  50 


I 

/ 


I 

s 

; 

; 

s 
s 
s 
s 


1   2  3  4  X  X  X 


1  2  3  A  X  X  X 


2  3  A  X  X  X 


1  2  3  A  5f  X  f 


Figure  1 — Mean  performance  of  five  measured  traits  of  cones  and  seeds  from  open  pollination_ 
in  four  populations  (1 ,  2,  3,  4)  compared  to  controlled  cross-pollination  (f )  and  self-pollination  (^) 
(x=  average  across  1 ,  2,  3,  4  populations). 


4.1  to  5.8  cm,  with  a  mean  of  4.9  ±  0.4  cm  for  open- 
pollinated; 

3.5  to  6.0  cm,  with  a  mean  of  4.8  ±  0.6  cm  for  cross- 
pollinated; 

3.7  to  5.5  cm,  with  a  mean  of  4.7  ±  0.6  cm  for  self- 
pollinated. 

Variation  coefficients  (VC)  for  open-pollinated,  cross- 
pollinated,  and  self-pollinated  were  7.9  percent,  12.8  per- 
cent, and  12.1  percent,  respectively. 


Table  4 — Comparison  between  statistical  parameters  of  cones  and  seeds  in 
the  Gemenele  population,  according  to  the  pollination  type 


Type  of   Parameters 


Trait 

pollination^ 

k±a 

VC 

0 

CL 

OP 

4.9 

± 

0.4 

0.15 

7.9 

4.1 

5.8 

CP 

4.8 

± 

0.6 

0.37 

12.8 

3.5 

6.0 

SP 

4.7 

± 

0.6 

0.36 

12.1 

3.7 

5.5 

CD 

OP 

4.2 

+ 

0.3 

0.11 

7.9 

3.6 

5.0 

CP 

4.3 

+ 

0.4 

0.15 

9.1 

3.7 

5.4 

SP 

4.2 

± 

0.4 

0.16 

9.4 

3.6 

4.9 

SC 

OP 

40.9 

+ 

9.6 

91.31 

23.3 

22.5 

60.2 

CP 

62.0 

± 

16.5 

272.24 

26.6 

41.4 

96.4 

SP 

64.0 

+ 

12.9 

167.12 

20.3 

44.9 

85.7 

sew 

OP 

9.8 

+ 

2.7 

7.50 

28.0 

4.6 

18.1 

CP 

14.1 

± 

4.4 

19.56 

31.3 

7.0 

24.5 

SP 

13.8 

+ 

3.1 

9.66 

22.5 

9.5 

17.6 

1,000-SW  OP 

238.0 

± 

47.4 

2,243.00 

19.9 

170.0 

352.0 

CP 

188.0 

± 

31.3 

978.00 

16.6 

11.7 

26.5 

SP 

179.0 

± 

25.6 

656.94 

14.3 

14.7 

22.3 

'OP  =  open  pollination;  CP  =  controlled  cross-pollination  in  a  10  x  10  diallel;  SP  =  self 
pollination  of  10  parents. 


Similar  data  can  be  foimd  for  cone  diameter,  seeds  per 
cone,  seeds  per  cone  weight,  and  1,000-seed  weight  in 
table  4. 

CORRELATIONS 

Significant  (p  <  0.05)  and  highly  significant  (p  <  0.01) 
correlations  were  found  between  cone  length  and  cone  di- 
ameter, cone  length  and  seeds  per  cone,  cone  length  and 
seeds  per  cone  weight;  cone  diameter  and  seeds  per  cone, 
cone  diameter  and  seeds  per  cone  weight;  seeds  per  cone 
and  seeds  per  cone  weight.  No  significant  correlation 
was  found  between  1,000-seed  weight  and  any  other  trait 
(table  5).  These  strong  correlations  among  the  main  cone 
and  seed  traits  suggest  that  improvement  (quantitatively) 
of  seed  production  could  be  made  by  indirect  selection;  for 
example,  selection  for  cone  length,  as  an  easily  measur- 
able trait,  will  cause  an  increase  in  seeds  per  cone  and 
seeds  per  cone  weight,  and  consequently  in  seed  production. 

TYPE  OF  DISTRIBUTION 

The  distribution  frequencies  of  cone  length,  cone  diam- 
eter, seeds  per  cone,  seeds  per  cone  weight,  and  1,000- 
seed  weight  were  very  close  to  normal  distribution  (fig.  2). 
According  to  genetic  theory  (Mather  and  Jinks  1977),  this 
pattern  of  distribution  is  specific  to  quantitative  traits. 
Such  traits  are  polygenically  controlled. 


156 


Table  5 — Phenotypic  correlations  among  Pinus  cembra  cone  and 
seed  traits  (Df  =  1 8) 


Trait 

CL 

CD 

SC 

sew 

1  nnn.sw 

CL 

1.000 

0.577- 

0.636- 

0.669- 

0.272 

CD 

1.000 

0.499* 

0.647- 

0.227 

SC 

1.000 

0.636- 

0.325 

sew 

1.000 

0.396 

1,000-SW 

1.000 

DISCUSSION 

Our  observations  in  the  main  Carpathian  populations 
indicated  that  tree,  locahty,  and  year  in  which  the  cones 
were  initiated  may  significantly  affect  cone  and  seed 
traits.  As  the  measurements  were  performed  on  cones 
collected  in  two  different  years,  the  comparisons  were 
perhaps  not  entirely  valid. 


The  seeds  per  cone  and  seeds  per  cone  weight  from  open- 
pollinated  cones  ranked  lower  than  the  same  parameters 
from  both  cross-pollinated  and  self-pollinated  cones;  but, 
surprisingly,  1,000-seed  weight  from  open-pollinated 
cones  ranked  higher  than  1,000-seed  weight  from  cross- 
pollinated  cones.  The  lack  of  significant  correlation 
between  1,000-seed  weight  and  seeds  per  cone  weight 
(table  5)  partially  explains  this  unexpected  resvilt.  Also, 
these  differences  could  be  attributable  to  the  biological 
and  climatic  factors  (temperature,  moistm-e,  and  wind) 
that  occurred  in  1990  and  1991. 

The  cone  length  and  cone  diameter  means  from  self- 
pollinated  cones  were  similar  to  cone  length  and  cone 
diameter  means  from  both  open-pollinated  and  cross- 
pollinated  cones,  but  the  seeds  per  cone  and  seeds  per 
cone  weight  means  from  self-pollinated  cones  were  similar 
to  seeds  per  cone  and  seeds  per  cone  weight  means  from 
cross-pollinated  cones  and  ranked  higher  them  open- 
pollinated  cones.  Consequently,  stone  pine  was  found 


£  20 


30 


^  20  - 


O 


£  10 


330 


3.7    3.9     4.2    4.5    4.8    5.1     5.4    5.7  6.0 

Cone  Length  (cm) 


3.1    3.4     3.7    4.0    4.3   4.6  4.9 

Cone  Diameter  (cm) 


^  20 


0> 


S>  10 


10    20     30     40     50     60     70     80  90 

Seeds  per  Cone  (No.) 


30  - 


^   20  ■ 


o 


2>  10 


4     7     10    13    16    19  22 


>g  20- 


0) 


0) 


150  190    230   270  310  350    390  430 


Seeds  per  Cone  Weight  (g)  1 ,000-Seed  Weight  (g) 

Figure  2— Frequency  distribution  in  the  Pinus  cembra  populations  evaluated  for  five  traits. 


157 


to  be  highly  self-fertile.  This  led  to  the  conclusion  that 
a  very  high  proportion  of  the  seeds  produced  by  wind  pol- 
lination would  be  selfs;  this  is  an  undesirable  characteris- 
tic since  selfed  seed  produces  slower  growing  trees  and 
the  seedlings  have  a  lower  survival  rate. 

Finally,  it  should  be  stressed  that  although  the  stone 
pine  cone  diameter,  cone  length,  and  seeds  per  cone 
means  from  the  Alps  exceeded  the  same  trait  means  from 
the  Carpathians,  the  1,000-cone  weight  mean  from  the 
Carpathians  was  similar  to  that  from  the  Alps. 

CONCLUSIONS 

Southern  populations  were  similar  in  cone  length  and 
cone  diameter  but  not  in  seeds  per  cone,  seeds  per  cone 
weight,  and  1,000-seed  weight.  Southern  populations 
ranked  high  and  the  northern  ones  ranked  low  in  all 
measured  traits. 

Very  high  variation  in  both  seeds  per  cone  and  seeds 
per  cone  weight  was  found  within  each  population,  but 
variation  in  1,000-seed  weight  was  moderate,  and  it  was 
low  in  cone  length  and  cone  diameter  traits. 

Cone  length  and  cone  diameter  from  open-pollinated, 
cross-pollinated,  and  self-pollinated  cones  were  similar, 
while  seeds  per  cone,  seeds  per  cone  weight,  and  1,000- 
seed  weight  were  not. 

The  strong  correlations  among  the  main  cone  and  seed 
traits  suggest  that  genetic  improvement  in  seed  produc- 
tion could  be  attained  by  indirect  selection. 

All  measured  traits  displayed  a  continuous  variation, 
suggesting  polygenic  control. 

ACKNOWLEDGMENTS 

Gratitude  is  expressed  to  Dr.  H.  Kriebel  from  Ohio 
State  University  for  the  revision  of  the  English  version 
of  the  paper  and  for  the  useful  suggestions. 

REFERENCES 

Beldie,  A.  1941.  Observatii  asupra  vegetatiei  lemnoase 
din  Muntii  Bucegi.  Analele  ICEF,  Seria  1,  vol.  6:  39-43. 

Blada,  I.  1982.  Relative  blister  rust  resistance  of  native 
and  introduced  white  pines  in  Romania.  In:  Heybroek,  H.; 
[and  others],  eds.  Resistance  to  diseases  and  pests  in 
forest  trees.  PUDOC,  Wageningen,  The  Netherlands: 
415-416. 

Blada,  I.  1987.  Genetic  resistance  to  Cronartium  ribicola 
and  height  growth  in  some  five  needle  pines  and  of 


some  inter-specific  hybrids.  Bucharest:  Academy  of 
Agricultural  and  Forestry  Sciences.  146  p.  Thesis. 

Blada,  I.  1990a.  Breeding  Pinus  cemhra  by  intra-  and 
inter-specific  crosses.  For.  Res.  Inst.,  Bucharest,  Re- 
search Plan,  No.  7/1990. 

Blada,  I.  1990b.  Genetic  variability  of  some  traits  in  two 
Pinus  cembra  natural  populations.  In:  Garrett,  P.  W., 
ed.  Proceedings  of  a  s3miposium  on  white  pines  prov- 
enances and  breeding.  Gen.  Tech.  Rep.  NE-155.  Radnor, 
PA:  U.S.  Department  of  Agriculture,  Forest  Service, 
Northeastern  Forest  Experiment  Station:  56-68. 

Blada,  I.  1992a.  Pinus  cembra  and  its  inter-specific  hy- 
brids. Paper  presented  at  the  lUFRO  Centennial  Meet- 
ing, Berlin-Eberswalde,  August  31-September  5,  1992. 

Blada,  I.  1992b.  Diallel  cross  in  Pinus  cembra.  I.  Variation 
in  cone  and  seed  traits.  [In  preparation]. 

Contini,  L.;  Lavarelo,  Y.  1982.  Le  pin  cembro:  repartition, 
ecologie,  sylviculture  et  production.  Paris:  INRA,  ISBN. 
197  p. 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  IVIisc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture,  For- 
est Service.  97  p. 

Griffing,  B.  1956.  Concept  of  general  and  specific  com- 
bining ability  in  relation  to  diallel  crossing  systems. 
Australian  J.  Sci.  9:  465-493. 

Gubesch,  L.  1971.  Raspandirea  relictului  glaciar  zambrul 
(Pinus  cembra)  pe  versantii  sudici  ai  unor  masive  din 
Calimani.  Ocrotirea  Naturii,  15,  2:  149-159. 

Holzer,  K.  1972.  Intrinsic  qualities  and  growth  potential 
of  Pinus  cembra  and  P.  peuce  in  Europe.  In:  Bingham, 
R.  T.;  [and  others],  eds.  Biology  of  rust  resistance  in  for- 
est trees.  Misc.  Publ.  Washington,  DC:  U.S.  Department 
of  Agriculture,  Forest  Service:  99-110. 

Holzer,  K.  1975.  Genetics  of  Pinus  cembra  L.  Anales 
Forestales,  Zagreb.  6/5:  139-158. 

Mather,  K.;  Jinks,  J.  L.  1977.  Introduction  to  biometrical 
genetics.  Chapman  and  Hall,  Ltd.  233  p. 

Moser,  L.  1960.  Verbreitung  und  Bedentung  der  Zirbe 
im  italienishcen  Alpengebiet.  Jb.  Ver.  Z.  Schutz 
Alpenpfl.u.-tiere.  25:  16-21. 

Oarcea,  Z.  1966.  Contributii  la  cunoasterea  raspandirii  si 
vegetarii  pinului  cembra  in  Retezat.  Revista  Pad.  9: 
495-497. 

Rohmeder,  E.;  Rohmeder,  M.  1955.  Untersuchungen  uber 
das  Samentragen  und  Keimen  der  Zirbelkiefer  (Pinus 
cembra)  in  den  Bayrischen  Alpen.  Allg.  Forstzeitschrift. 
10:  83. 

Tataranu,  I.  D.;  Costea  C.  1952.  Un  arbore  de  interes 
forestier:  Pinus  cembra.  Revista  Padurilor.  11:  3-14. 


158 


ROLE  OF  NUTCRACKERS  ON  SEED 
DISPERSAL  AND  ESTABLISHMENT  OF 
PINUS  PUMILA  AND  P.  PENTAPHYLLA 

Mitsuhiro  Hayashida 


AbBtract— Nutcrackers  (Nucifraga  caryocatactes)  harvest  almost 
all  Pinus  pumila  Kegel  cones  and  cache  the  seeds  in  the  soil.  Most 
P.  pentaphylla  'M.ayr  seeds  scatter  after  the  cones  open,  while  nut- 
crackers collect  the  seeds  still  held  in  the  cones  and  then  cache  the 
seeds.  Many  seedlings  of  P.  pumila  and  P.  pentaphylla  are  often 
found  in  clusters  consisting  of  several  individuals  of  the  same  age. 
Almost  all  P.  pumila  seedlings  and  clusters  of  P.  pentaphylla 
seedlings  on  barren  slopes  likely  originate  from  seeds  cached  by 
nutcrackers. 


Stone  pines  (subsection  Cembrae  in  Pinus)  have  large 
wingless  seeds  that  are  dispersed  by  caching  of  vertebrates 
(Hayashida  1989a;  Hutchins  and  Lanner  1982;  Mattes 
1982;  Tomback  1982).  Nutcrackers  {Nucifraga  spp.)  are 
the  main  dispersal  agents  of  seeds  of  these  pines  (Lanner 
1989).  The  nutcracker  also  caches  other  large  wingless 
seeds  or  short-winged  seeds  in  section  Strobus  (Lanner 
and  Vander  Wall  1980;  Tomback  1990;  Vander  Wall  and 
Balda  1977). 

Pinus  pumila  Regel  (Japanese  stone  pine)  is  a  dwarf  pine. 
Its  range  extends  through  Japan  and  Korea  into  Siberia, 
Kamchatka  (Critchfield  and  Little  1966).  On  most  of  the 
high  mountains  in  Japan,  a  vegetational  zone  dominated 
by  P.  pumila  occurs  above  the  forest  limit.  This  area  is 
called  the  "Pinus  pumila  zone"  by  Japanese  ecologists.  The 
regeneration  of  P.  pumila  thickets  is  mainly  due  to  a  veg- 
etative regeneration  by  adventitious  roots  (Okitsu  and  Ito 
1983).  Large  wingless  seeds  of  P.  pumila  are  known  to  be 
dispersed  by  nutcrackers  (Saito  1983;  Turcek  and  Kelso 
1964),  but  a  detailed  observation  for  seed  dispersal  and 
caching  behavior  is  needed. 

Pinus  parviflora  Sieb.  et  Zucc.  (Japanese  white  pine)  is 
distributed  throughout  the  Japanese  islands  (Critchfield 
and  Little  1966).  This  species  consists  of  two  geographical 
varieties,  which  intergrade  in  central  Japan  (Hayashi  1954). 
The  northern  variety  is  sometimes  called  P.  pentaphylla 
Mayr;  the  southern  P.  himekomatsu  Miyabe  and  Kudo  or 
a  variety  of  the  former.  Since  the  seed  wings  of  P.  penta- 
phylla are  longer  than  those  of  P.  himekomatsu  (Ishii  1968), 
I  distinguished  between  the  two  pines.  There  are  few  eco- 
logical studies  of  these  pines  except  for  community  struc- 
ture (Tatewaki  and  others  1960;  Yoshioka  and  Saito  1962). 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Mitsuhiro  Hayashida  is  Associate  Professor,  Department  of  Bioenviron- 
ment,  Faculty  of  Agriculture,  Yamagata  University,  Wakaba,  Tsuruoka 
997,  Japan. 


Pinus  pumila  and  P.  pentaphylla  are  in  the  subgenus 
Strobus  and  are  so  similar  that  they  naturally  hybridize 
(P.  hakkodensis  Makino)  (Ishii  1968);  but  these  pines  differ 
in  seed,  cone,  and  other  characteristics.  The  objectives  of 
this  paper  are  to  describe  the  processes  from  seed  dispersal 
to  seedling  estabUshment  of  P.  pumila  and  P.  pentaphylla 
and  to  examine  the  role  of  nutcrackers  in  their  regeneration. 

STUDY  AREA 

The  study  was  conducted  on  Mount  Apoi  (811  m  above  sea 
level),  which  is  located  in  south-central  Hokkaido,  Japan 
(42°6'  N.,  143°2'  E.).  It  is  composed  of  ultramafic  rocks 
(dunite,  IherzoHte,  etc.)  (Niida  1984),  and  is  covered  with 
coniferous  forests.  The  P.  pumila  zone  extends  from  eleva- 
tions of  500  m  to  800  m.  Below  500  m,  P.  pentaphylla  for- 
est occupies  the  rocky  slopes  and  ridges,  while  the  greater 
part  of  the  mountain  is  covered  by  coniferous  forests  domi- 
nated by  Picea  glehnii,  Pinus  pentaphylla,  and  Abies 
sachalinensis  (Hayashida  1989b). 

METHODS 

The  number  and  distribution  of  2-year  cones  were  mapped 
to  determine  rates  of  seed  predation  by  vertebrates  and  in- 
sects. These  cones  were  counted  from  July  to  October  in 
1984-87.  The  observations  on  seed  harvesting,  transport- 
ing, caching,  and  retrieving  behavior  in  vertebrates  were 
made  in  1984-87  using  binoculars  and  a  telescope. 

Many  seedlings  of  P.  pumila  and  P.  pentaphylla  were 
found  on  barren  slopes  along  mountain  trails.  These  seed- 
lings were  often  found  in  clusters,  which  consisted  of  sev- 
eral seedhngs  of  the  same  age.  To  obtain  the  frequency 
distribution  of  number  of  individuals  per  cluster  in  both 
pines,  all  seedlings  on  barren  slopes  along  mountain  trails 
were  recorded. 

SEED  AND  CONE  MORPHOLOGY 

Pinus  pumila  has  wingless  seeds;  P.  pentaphylla  seeds 
have  wings  that  are  seed  length  (table  1).  Seeds  of  both 
pines  have  the  same  thick  coats.  Pinus  pumila  seeds  are 
significantly  heavier  than  those  of  P.  pentaphylla,  though 
the  seeds  of  P.  pentaphylla  are  larger. 

Pinus  pumila  cones  do  not  open  when  they  ripen,  but 
P.  pentaphylla  cones  open  at  maturity  and  release  seeds 
to  fall  free.  Pinus  pumila  cones  weigh  less  than  P.  penta- 
phylla cones,  but  contain  more  seeds. 


159 


Table  1 — Seed  and  cone  characteristics  of  Pinuspumila  and  P.  pentaphylla 


Pinus  pumila    P.  pentaphylla 


MAAQIirOITIAnt 

d  1  Id  It 

Moan  +  Qn 

M 

Moon  4-  QD 

Kl 

n 

P< 

Seed  length  (mm) 

7.700  ±  0.600 

105 

10.000  ±  0.700 

105 

0.001 

Dry  weight  (g) 

0.100  ±  0.014 

36 

0.068  ±  0.014 

20 

0.001 

Seed  coat  thickness  (mm) 

0.390  ±  0.040 

20 

0.390  ±  0.040 

20 

NS 

Length  of  wings  (mm) 

wingless 

10.500  +  1.600 

105 

Cone  mass  (g) 

11.00013.300 

32 

24.800  ±  7.000 

27 

0.001 

Number  of  seeds  per  cone 

42.800  ±12.900 

45 

28.900  ±9.100 

38 

0.001 

Dehiscence 

not  open 

open 

'Mann-Whitney  U  test. 


SEED  DISPERSAL  OF  P.  PUMILA 

Figure  1  shows  the  seasonal  disappearance  of  2-year 
cones  of  P.  pumila  in  1985-88.  All  cones  disappeared  from 
trees  by  mid-October  every  year.  I  observed  nutcrackers 
(Nucifraga  caryocatactes),  squirrels  (Sciurus  vulgaris  ori- 
entis),  and  chipmunks  (Tamias  sibiricus  lineatus)  harvest- 
ing cones.  Five  species  of  vertebrates  harvested  and  car- 
ried seeds  to  caches:  nutcracker,  vsiried  tit  (Parus  varius), 
nuthatch  (Sitta  europaea),  squirrel,  and  chipmunk.  The 
nutcracker  was  the  most  frequently  observed  vertebrate 
carr3dng  P.  pumila  seeds. 

Nutcrackers  usually  harvest  seeds  by  removing  the 
cones  and  flying  with  them  to  perches  where  they  extract 
the  seeds  and  hold  them  in  their  sublingual  pouches.  Var- 
ied tits  and  nuthatches  were  able  to  take  P.  pumila  seeds 
exposed  by  nutcracker  foraging  and  eat  them.  They  were 
seen  caching  pine  seeds  on  a  tree  or  in  the  soil.  Squirrels 
visited  the  P.  pumila  scrub  from  a  coniferous  forest  to  for- 
age on  pine  seeds.  They  often  carried  a  whole  cone  to  cache 
it  in  scatter-hoarding.  Chipmunks  harvest  pine  seeds  in 
their  cheek  pouches.  They  carried  about  50  seeds  on  a  trip 
and  cached  them  in  the  soil. 

These  results  indicate  that  nutcracker  is  the  main  seed 
disperser  of  P.  pumila. 


(0 

E 
« 
oc 


Jul. 

Aug. 

Sep. 

Oct. 

Figure  1 — Retention  of  mature  cones  of  Pinus  pumila 
on  trees  (in  years  1985-88)  as  related  to  season. 


SEED  DISPERSAL  OF 
P.  PENTAPHYLLA 

Cones  taken  by  vertebrates  or  damsiged  by  insects 
(cone-boring  lepidopteran  larvae)  accounted  for  less  than 
20  percent  of  the  total.  Usually,  nearly  90  percent  of  the 
cones  open  on  the  trees  from  mid-September  through  early 
October.  After  opening,  about  80  percent  of  the  seeds  im- 
mediately scatter  to  the  ground,  due  to  the  fact  that  77 
percent  of  P.  pentaphylla  cones  were  oriented  downward 
(Hayashida  1989b). 

Nutcrackers,  varied  tits,  and  nuthatches  harvested  and 
cached  seeds  still  held  in  the  upward-oriented  open  cones. 
In  the  years  when  P.  pumila  crop  sizes  were  small,  nut- 
crackers harvested  P.  pentaphylla  seeds  more  frequently. 
They  extracted  the  several  seeds  from  an  open  cone,  peeled 
off  the  seed  wings,  and  put  seeds  into  their  sublingual 
pouches.  They  moved  rapidly  around  a  tree  and  went  from 
tree  to  tree. 

NUTCRACKER  CACHING  BEHAVIOR 

Nutcrackers  began  to  cache  P.  pumila  seeds  in  early 
August  when  the  seeds  were  mature,  and  they  continued 
to  harvest  seeds  until  the  seed  crop  was  depleted.  They 
cached  the  seeds  in  the  soil  at  a  depth  of  2  to  3  cm.  Cache 
size  (the  number  of  seeds  per  cache)  ranged  from  1  to  51. 
Nutcrackers  cached  P.  pumila  seeds  mainly  in  coniferous 
forests. 

In  mid-September,  P.  pentaphylla  cones  began  to  open, 
then  nutcrackers  began  to  harvest  seeds  from  the  opening 
cones.  Nutcrackers  scatter-hoarded  almost  all  seeds  in  the 
soil  (97  percent)  except  for  two  caches  containing  one  and 
two  seeds  that  were  established  in  the  bark  on  fallen  trees. 
Cache  size  ranged  from  1  to  40  seeds.  Seventy-eight  caches 
were  recorded  in  coniferous  forests  and  only  one  cache  in 
the  Pinus  pumila  zone. 

CLUMPS  OF  SEEDLINGS 

Eighty-four  percent  of  the  occurrence  of  P.  pumila  seed- 
lings were  in  groups  of  two  or  more  trees  (table  2).  These 
seedlings  were  1  km  in  distance  from  the  nearest  seed  tree. 


160 


Table  2 — Number  of  individuals  per  clump  of  Pinus  pumila  and  P.  pentaphylla  on 
barren  slopes  along  mountain  trails 


Number  of  clumoQ 

Numbor  of  inHiviclijalQ 

(percent) 

per  clump 

Species 

Single       Clump  (2<) 

Maximum       Mean  ±  SD 

Pinus  pumila 

9(16)  47(84) 

26              5.6  ±5.3 

P.  pentaphylla 

137(72)  54(28) 

25              2.6  ±  3.8 

Twenty-eight  percent  of  P.  pentaphylla  seedlings  were 
growing  in  groups  of  two  or  more  trees.  There  were  many 
seed  trees  of  P.  pentaphylla  along  the  moimtain  trails.  Sin- 
gle seedlings  might  originate  from  naturally  scattered  fallen 
seeds,  and  the  seedlings  in  cliunps  probably  originated  from 
caches  by  nutcrackers. 

DISCUSSION 

All  mature  cones  of  P.  pumila  are  harvested  by  verte- 
brates every  year  (fig.  1).  My  observations  indicate  that 
almost  all  of  the  transported  seeds  were  carried  by  nut- 
crackers. Nutcrackers  cached  seeds  in  the  soil  at  a  depth 
of  2  to  3  cm.  It  is  shallow  enough  to  permit  seedling  estab- 
lishment. Therefore,  seed  dispersal  of  P.  pumila  is  heavily 
dependent  on  seed  caching  by  nutcrackers. 

Nutcrackers  carried  and  cached  P.  pumila  seeds  mainly 
into  coniferous  forests.  In  general,  P.  pumila  seedhngs  can- 
not grow  and  establish  in  coniferous  forests  even  if  cached 
seeds  survive  and  germinate.  However,  several  P.  pumila 
trees  were  found  that  developed  and  matured  on  the  rocky 
ridges  or  barren  slopes  in  coniferous  forests.  These  facts 
may  indicate  that  the  nutcracker  is  not  an  efficient  seed 
disperser  of  P.  pumila  under  the  present  conditions.  Pinus 
pumila  adapted  its  specialized  tree  form  and  other  features 
to  the  habitat  that  is  characterized  by  strong  winds  and 
heavy  snow  accumulation  in  winter.  Thus,  the  P.  pumila 
zone  is  developed  on  deforested  areas  in  the  boreal  subal- 
pine  belt.  In  Europe,  nutcrackers  cached  P.  cembra  seeds 
in  coniferous  stands  and  sometimes  cached  them  above 
timberline  (Mattes  1982).  Seed  dispersal  over  wide  areas 
by  nutcrackers  probably  played  an  important  role  in  ex- 
panding the  range  of  P.  pumila. 

Pinus  pentaphylla  has  dehiscent  cones  and  winged  seeds. 
Most  of  the  seeds  scattered  when  cones  opened.  Nutcrack- 
ers often  harvested  and  cached  seeds  still  held  in  the  up- 
ward open  cones.  About  90  percent  of  P.  pentaphylla  trees 
found  in  coniferous  forests  were  individual  single  trees,  but 
on  the  ecotone  between  coniferous  forest  and  the  Pinus 
pumila  zone,  half  of  the  occurrences  of  trees  were  in  clumps 
of  two  or  more  individuals  (Hayashida  1989b).  About  30 
percent  of  the  seedlings  were  growing  in  clumps  on  barren 
slopes  in  the  coniferous  forest  zone.  These  seedlings  in 
climips  probably  originated  from  seeds  cached  by  nutcrack- 
ers. Seed  dispersal  by  nutcrackers  enables  P.  pentaphylla 
to  invade  areas  that  are  inaccessible  to  other  coniferous 
trees. 

The  northern  variety,  P.  pentaphylla,  has  longer  seed 
wings  than  the  southern  variety,  P.  himekomatsu  (Ishii 
1968).  It  is  possible  that  these  two  varieties  have  different 


morphological  features  and  seed-dispersal  syndromes.  If 
so,  why?  I  would  study  these  problems  with  an  ecological 
approach. 

ACKNOWLEDGMENTS 

I  thank  Urakawa  Local  Forest  Management  Office, 
Hokkaido,  and  Oji  Forestry  and  Landscaping  Co.  for  per- 
mission to  work  in  the  area  and  for  information  on  the 
area. 

REFERENCES 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  distri- 
bution of  pine  of  the  world.  Misc.  Publ.  991.  Washington, 
DC:  U.S.  Department  of  Agriculture,  Forest  Service.  97  p. 

Hayashi,  Y.  1954.  The  natural  distribution  of  important 
trees,  indigenous  to  Japan.  Conifers  Rep.  3.  Bulletin  of 
the  Government  Forestry  Experiment  Station.  75:  1-173. 
[In  Japanese  with  English  svmimary]. 

Hayashida,  M.  1989a.  Seed  dispersal  by  red  squirrels  and 
subsequent  establishment  of  Korean  pine.  Forest  Ecology 
and  Management.  28:  115-129. 

Hayashida,  M.  1989b.  Seed  dispersal  and  regeneration 
pattern  of  Pinus  parviflora  var.  pentaphylla  on  Mt.  Apoi 
in  Hokkaido.  Research  Bulletins  of  the  College  Experi- 
ment Forests,  Facvdty  of  Agriculture,  Hokkaido  Univer- 
sity. 46(1):  177-190.  [In  Japanese  with  English  stmimary]. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Ishii,  S.  1968.  The  basic  forestry  study  on  genus  Pinus, 
especially  its  taxonomic  and  geographical  consideration. 
Journal  of  the  Faculty  of  Agriculture,  Kochi  University. 
19:  1-114.  [In  Japanese  with  English  svunmary]. 

Lanner,  R.  M.  1989.  Biology,  taxonomy,  evolution,  and 
geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposiimi  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mouintain  resource;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain  Research  Station:  14-24. 

Lanner,  R.  M.;  Vander  Wall,  S.  B.  1980.  Dispersal  of  lim- 
ber pine  seed  by  Clark's  nutcracker.  Journal  of  Forestry. 
78:  637-639. 

Mattes,  H.  1982.  Die  Lebensbemeinschaft  von  Tannenhaher, 
Nucifraga  caryocatactes  (L.),  und  Arve,  Pinus  cembra  L., 
und  ihre  forstliche  Bedeutiing  in  der  oberen  Gebirgs- 
waltstufe,  Berichte.  241.  74  p. 


161 


Niida,  K.  1984.  Petrology  of  the  Horoman  utramafic  rocks 
in  the  Hidaka  Metamorphic  Belt,  Hokkaido,  Japan.  Jour- 
nal of  the  Faculty  of  Science,  Hokkaido  University,  Ser. 
IV.  21:  197-250. 

Okitsu,  S.;  Ito,  K.  1984.  Vegetation  d3niamics  of  the  Siberian 
dwarf  pine  (Pinus  pumila  Kegel)  in  the  Taisetsu  moun- 
tain range  Hokkaido,  Japan.  Vegetatio.  58:  105-113. 

Saito,  S.  1983.  Caching  of  Japanese  stone  pine  seeds  by 
nutcrackers  at  the  Shiretoko  Peninsula,  Hokkaido.  Tori. 
32:  13-20.  [In  Japanese  with  English  summary]. 

Tatewaki,  M.;  Tsujii,  T.;  Kawano,  S.  1960.  The  community 
structure  and  distribution  of  Pinus  parvifLora  forest  in 
Hokkaido.  Japanese  Journal  of  Ecology.  10:  120-123.  [In 
Japanese  with  English  summary]. 


Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds  by 
Clark's  nutcracker:  a  mutualism  hypothesis.  Journal  of 
Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.  1990.  The  evolution  of  bird-dispersed 
pines.  Evolutionary  Ecology.  4:  185-219. 

Turcek,  F.  J.;  Kelso,  L.  1968.  Ecological  aspects  of  food 
transportation  and  storage  in  the  Corvidae.  Communi- 
cations in  Behavioral  Biography,  Al:  277-297. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of  the 
Clark's  nutcracker  and  the  pinon  pine  for  efficient  seed 
harvest  £md  dispersal.  Ecological  Monographs.  47:  89-111. 

Yoshioka,  K;  Saito,  K.  1962.  Differences  in  the  distribution 
of  forest  communities  due  to  topography  of  the  habitats 
in  Okunikkawa,  Miyagi  Prefecture.  Ecological  Review. 
15:  213-220. 


162 


ROLE  OF  VARIOUS  ANIMALS  IN 
DISPERSAL  AND  ESTABLISHMENT 
OF  WHITEBARK  PINE  IN  THE  ROCKY 
MOUNTAINS,  U.S.A. 

Harry  E.  Hutchins 


Abstract — The  dispersal  and  establishment  of  whitebark  pine 
has  been  attributed  to  Clark's  nutcracker.  But  what  is  the  role 
of  other  animals  who  forage  on  whitebark  pine  seed?  This  study 
looks  at  the  role  of  the  diurnal  birds  and  mammals  that  forage 
on  whitebark  pine  cones  and  seed  in  western  Wyoming.  A  total 
of  1,005  cones  were  carefully  observed  in  both  contiguous  "forest" 
stands  and  open-grown  "meadow"  trees  in  early  November.  Red 
squirrels  harvested  most  of  the  whitebark  pine  seed  in  forested 
plots  while  Clark's  nutcracker  removed  most  of  the  seed  from  the 
meadow  trees.  Only  a  small  percentage  of  seed  went  to  ground 
squirrels,  chipmunks,  Stellar's  jays,  ravens,  etc.  The  foraging 
behavior  as  well  as  the  importance  of  these  animal  species  to 
whitebark  pine  regeneration  is  discussed. 


Many  animal  species  have  been  observed  foraging  on 
the  nutritious  seeds  of  whitebark  pine  (Pinus  albicaulis). 
But  exactly  which  of  these  animals  plays  a  role  in  seed- 
ling establishment  and  to  what  degree?  Which  animals 
are  seed  predators?  Can  whitebark  pine  seed  germinate 
from  that  seed  which  is  dropped  through  animal  foraging 
accidents?  Do  animals  harvest  all  the  seed  in  the  fall  or 
can  cones  fall  to  the  ground  and  seed  germinate  as  the 
cone  decays?  This  paper  will  address  these  questions. 

The  role  of  various  animals  in  the  dispersal  and  estab- 
lishment of  whitebark  pine  was  examined  by  observing 
1,005  cones  and  determining  where  and  how  the  seeds 
were  dispersed.  By  primarily  observing  the  behavior  of 
diurnal  birds  and  mammals  in  the  subalpine  canopy,  I 
was  able  to  determine  the  degree  of  influence  each  species 
has  on  seed  dispersal  and  seedling  establishment. 

METHODS 

The  study  was  primarily  conducted  at  the  Squaw  Basin- 
Togwotee  Pass  area  (Bridger-Teton  National  Forest;  fig.  1). 
This  area  of  high-elevation  meadows  offered  distant  views 
of  bird  activity,  whitebark  pine  growing  in  both  contigu- 
ous forest  and  open-grown  situations,  and  an  abundant 
cone  crop  for  whitebark  pine  diuing  1980.  Besides  white- 
bark pine,  the  forest  stands  were  composed  of  Engelmann 
spruce  (Picea  engelmannii),  subalpine  fir  {Abies  lasio- 
carpa),  and  a  minor  amoimt  of  lodgepole  pine  (Pinus  con- 
torta).  The  moraine  meadow  ridges  were  pioneered  by 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Harry  E.  Hutchins  is  Natural  Resource  Instructor,  Itasca  Community 
College,  Grand  Rapids,  MN. 


lone  whitebark  pine.  This  provided  an  opportimity  to 
study  two  different  types  of  commimities — "forest"  and 
"meadow"  (fig.  2). 

Observations  of  animal  activity  were  also  made  at 
Mount  Washburn  (2,680  to  3,140  m,  Yellowstone  National 
Park,  WY)  and  Surprise  Lake  (2,960  m.  Grand  Teton  Na- 
tional Park,  WY).  These  areas  were  similar  to  the  Squaw 
Basin  site  except  they  lacked  open-grown,  cone-bearing 
trees.  Details  on  the  study  sites  can  be  foxmd  in  Hutchins 
and  Lanner  (1982)  and  Vander  Wall  and  Hutchins  (1983). 

Several  cone-bearing  whitebark  pines  were  chosen  for 
observation  in  both  continuous  forest  in  which  red  squir- 
rels {Tamiasciurus  hudsonicus)  were  present  and  in  the 
open  meadow,  at  least  several  hundred  meters  from  the 
nearest  forest  edge  (these  sites  lacked  squirrels).  In  each 
of  these  two  types  of  sites,  1,005  whitebark  pine  cones 
were  observed  during  the  period  Jvdy  3  to  November  2, 
1980.  The  cones  were  scattered  among  several  mature 
trees  and  were  counted  at  1-  to  2-week  intervals  using 
a  15-25  X  60-mm  spotting  scope.  Coimts  were  made  by 
standing  in  a  specific  marked  location  and  "mapping"  the 
cones  on  clear  acetate.  Changes  in  the  "cone  map"  at  each 
observation  were  recorded,  including  the  partial  removal 
of  a  cone.  Cone  count  data  were  converted  into  seeds  us- 
ing an  empirically  derived  value  for  the  mean  number  of 
seeds  per  cone  (50.4  seeds/cone,  Hutchins  1982).  Partially 
consimied  cones  were  tallied  by  estimating  from  the  ground 
the  percentage  of  seed  remaining  as  described  in  Hutchins 
and  Lanner  (1982).  The  seed  harvest  data  were  then  plot- 
ted against  ciunulative  time  for  both  forest  and  meadow 
sites.  At  each  observation  date  samples  of  seeds  were  col- 
lected to  determine  maturity  and  condition.  The  numbers 
of  filled,  discolored,  insect-attacked,  and  second-year- 
aborted  ovules  were  tallied.  Mean  dry  weight  of  shelled 
seeds,  seed  coat  thickness,  and  caloric  content  of  shelled 
seeds  were  all  obtained.  Germination  tests  were  also  con- 
ducted with  seeds  collected  at  10  collection  dates  during 
the  1980  field  season. 

Seedfall  (caused  by  animal  foraging)  below  the  tree  was 
estimated  by  placing  five,  1-m^  vnre  mesh  seed  traps  ran- 
domly below  trees.  The  tops  of  the  seed  traps  were  de- 
signed to  let  seed  fall  through  but  to  exclude  rodents. 
Data  on  the  nimiber  of  cones  in  the  tree,  the  area  of  the 
tree  crown,  and  the  number  of  seeds  falling  per  square 
meter  were  used  to  estimate  the  magnitude  of  seed  fall. 

Predation  on  seed  caches  of  whitebark  pine  was  also 
studied  in  1979  and  1980.  I  simulated  three  types  of 
caches:  (1)  seed  that  lands  on  the  soil  surface  from 
foraging  accidents;  (2)  seed  cached  at  a  depth  of  3  cm 


163 


MONTANA 


WYOMING 


GRAND  TEfoN 
NAT.'  PARk 

I    i  ® 

Surprise  Lake  ^ 

i  A 


SHOSHONE 
NATIONAL 
FOREST 


LEGEND 


—  PAVED  ROAD 
®      STUDY  SITE 

KILOMETERS 


WYOMING 


simulating  a  Clark's  nutcracker  (Nucifraga  columbiana) 
cache;  and  (3)  seed  cached  7  cm  deep,  which  simulates 
the  most  shallow  red  squirrel  seed  cache.  Each  individual 
cache  contained  10  seeds. 

Animal  time  budget  data  were  collected  for  the  diurnal 
species  found  foraging  in  the  whitebark  pine  ecosystem. 
Both  quantitative  and  qualitative  observations  were  made 
on  the  various  activities  and  behaviors  of  the  animals;  the 
methods  are  detailed  in  Hutchins  and  Lanner  (1982). 

SEED  DEVELOPMENT 

Hutchins  and  Lanner  (1982)  monitored  seed  develop- 
ment and  found  mean  seed  coat  thickness  and  seed  weight 
to  be  significantly  greater  during  the  August  31  to 
November  2  collecting  period  than  prior  to  these  dates. 
Clark's  nutcrackers  were  unable  to  extract  whole  seed 
from  a  cone  ;intil  August  13  due  to  the  thin,  fragile  seed 
coats.  The  characteristic  shell  fragments  of  an  unripened 
"nutcracker  cone"  were  left  as  evidence  of  their  foraging 
attempts.  Thus,  nutcrackers  were  unable  to  cache  whole 
seed  until  after  this  date. 

By  September  7-10,  the  cones  had  dried  and  turned  a 
dull  brown  from  their  previous  moist,  pitch-filled,  purple 
color.  Whitebark  pine  cones  are  often  referred  to  as  inde- 
hiscent;  however,  about  25  percent  of  the  cones  collected 
after  September  7  parted  their  scales  slightly  (4-8  mm; 
n  =  141).  This  still  was  not  enough  of  an  opening  to  allow 
the  seeds  to  fall  out  on  their  own  accord. 


Figure  1 — Locations  of  three  study  sites  in  Wyoming. 


164 


SEED  CROP  DEPLETION 

Figure  3  (from  Hutchins  and  Lanner  1982)  shows  seed 
crop  depletion  followed  a  logistic  curve  (r^  =  0.99  in  the 
forest  and    =  0.96  in  the  meadow).  Seed  harvesting  be- 
gan somewhat  earlier  in  the  forest  than  in  the  meadow. 
In  fact,  about  50  percent  of  the  seed  crop  had  been  har- 
vested in  the  forest  by  August  31,  while  in  the  meadow 
this  point  was  not  attained  until  September  25.  As  seen 
on  the  graph,  no  seeds  remained  from  this  mast  year  in 
the  forest  stands  by  November  1.  Only  0.1  percent  of  the 
seeds  still  resided  in  the  tree  crowns  of  the  meadow  trees, 
and  these  were  no  longer  there  when  checked  on  27  June, 
1981. 

Unless  aided  by  an  animal,  cones  of  whitebark  pine  do 
not  fall  off  a  tree  to  the  ground  except  rarely  during  ex- 
tremely large  cone  crops — as  occurred  in  1989  (Hutchins 
and  Lanner  1982;  Lanner  1982;  Mattson  and  Reinhart, 
these  proceedings).  The  following  incident  shows  the  ex- 
tent Clark's  nutcracker  will  go  to  acquire  seed.  Hutchins 
and  Lanner  (1982)  placed  double-layer  wire  hardware 
screening  over  several  cones  so  there  would  be  cones  to 
collect  for  germination  studies.  By  late  October — after 
available  whitebark  pine  seeds  had  become  scarce — the 
nutcrackers  had  ripped  open  the  screening  and  harvested 
the  seed.  Seed  that  does  drop  to  the  ground  from  foraging 
accidents  made  up  only  4.2  percent  of  the  seed  crop  and 
most  of  this  seed  (69  percent)  was  determined  unviable 
through  examination  of  the  contents  in  the  shell  during 
1980  (Hutchins  1990). 

ANIMAL  INTERACTIONS  WITH  SEED 

A  variety  of  diurnal  animals  were  observed  active  in  or 
imder  whitebark  pine  trees.  Many  of  these,  however,  were 
never  observed  foraging  on  whitebark  pine  seed  in  this 
study.  These  include:  gray  jay  {Perisoreus  canadensis). 


common  flicker  (Colaptus  auratus),  Cassin's  finch  (Cora- 
podacus  cassinii),  rosy  finch  (Leucosticte  sp.),  pine  siskin 
(Spinas  pinus),  dark-eyed  junco  (Junco  hymelis),  black 
billed  magpie  (Pica  pica),  pine  marten  (Martes  americana), 
coyote  (Canis  latrans),  and  weasel  (Mustela  sp.). 

Many  feces  of  the  mammal  species  were  examined,  but 
no  evidence  of  whitebark  pine  seeds  was  found.  Gray  jays 
were  commonly  found  in  the  crowns  of  whitebark  pine  but 
were  only  observed  hawking  insects  and  caching  fresh 
carrion  or  boli  in  pine  branches.  Dow  (1965)  found  this 
species  has  little  interest  in  pine  seed  during  feeding  tri- 
als, although  Turcek  and  Kelso  (1968)  describe  its  Eur- 
asian cousin,  the  European  jay  (P.  infaustus),  as  having 
been  observed  taking  and  caching  Siberian  stone  pine 
(Pinus  sibirica). 

Of  the  birds,  only  the  magpie  could  potentially  be  in- 
volved in  the  dispersal  of  whitebark  pine  (Smith  and  Balda 
1979).  It  is,  however,  very  rare  in  the  subalpine  zone  dur- 
ing the  maturation  of  whitebark  pine  seed  (Hutchins  and 
Lanner  1982);  consequently,  any  potential  harvesting  of 
seed  would  be  unimportant  to  the  regeneration  of  white- 
bark pine. 

Many  species  have  been  observed  foraging  on  whitebark 
pine  cones  (Hutchins  1990,  table  2).  At  this  point  I  would 
like  to  take  a  closer  look  at  their  role  in  the  dispersal  and 
establishment  of  this  pine. 

BIRD  DISPERSERS 

Clark's  Nutcracker — Nutcrackers  were  the  most 
common  resident  bird  to  visit  the  whitebark  pine  trees 
(Hutchins  and  Lanner  1982).  They  were  scattered  about 
whitebark  pine  stands  in  loose  flocks  foraging  and  caching 
seed  of  this  tree.  They  are  dependent  on  these  caches 
year  round  (Giuntoli  and  Mewaldt  1978;  Vander  Wall 
and  Balda  1977;  Vander  Wall  and  Hutchins  1983). 


August 


September 


October 


November 


Figure  3 — Seasonal  course  of 
whitebark  pine  seed  harvest  by 
vertebrates  in  Squaw  Basin, 
WY,  in  1980. 


165 


Nutcrackers  were  observed  harvesting  seed  as  early 
as  July  13,  1979.  During  July,  however,  they  appeared  to 
only  be  testing  the  cones  for  ripeness  and  primarily  feed- 
ing on  the  previous  year's  caches  until  mid-August.  Seed 
harvested  at  this  time  were  lost  to  regeneration  because 
the  seed  coats  were  broken  and  the  seed  is  ungerminable 
vintil  August  13.  In  1980,  the  birds  were  able  to  success- 
fully harvest  whole,  developed  seed  by  August  15.  Nut- 
crackers harvested  seed  from  cones  at  increasingly  faster 
rates  (Hutchins  and  Lanner  1982)  through  early  October, 
when  seed  became  hard  to  find  (fig.  3).  Nutcrackers  were 
never  seen  attempting  to  harvest  Engelmann  spruce  or 
lodgepole  pine,  even  as  the  whitebark  pine  seed  crop  dwin- 
dled in  October.  As  the  seed  supply  became  scarce  in  mid- 
October,  I  observed  one  bird  checking  over  50  cones  for  a 
period  of  613  seconds  without  finding  a  single  seed. 

By  November  2,  it  appears  nutcrackers  were  almost 
totally  dependent  on  their  new  seed  caches,  and  would 
remain  so  imtil  the  following  August.  These  caches  were 
placed  just  below  the  soil  surface  (2-3  cm),  and  over  70  per- 
cent of  the  caches  were  either  one-  two-  or  three-seed 
caches  (x  =  3.2  +  2.8  seeds/cache;  fig.  4).  The  largest  cache 
was  14  seeds.  These  data  agree  closely  with  Tomback 
(1978)  and  Vander  Wall  and  Balda  (1977),  although  the 
number  of  single-seeded  caches  was  higher. 

Clark's  nutcrackers  cached  their  seeds  in  almost  every 
conceivable  soil  type  and  microsite  in  the  various  study 
sites.  Specific  examples  were  given  in  Hutchins  (1990). 
Although  they  may  cache  seeds  on  a  wide  variety  of  sites 
in  the  Rocky  Mountains,  several  studies  have  indicated 
that  south-facing  slopes  seem  to  be  preferred  (Lanner 
1982;  Lanner  and  Vander  Wall  1980;  Snethen  1980; 
Tomback  1978;  Vander  Wall  and  Balda  1977).  Seedling 
establishment  appears  to  be  much  more  common  on  more 
moist  sites  in  the  Rocky  Mountains  (Arno  1986;  Arno  and 
Hoff  1989;  Vander  Wall  and  Hutchins  1983)  than  in  the 
Sierra  Nevada  (Tomback  1982). 


40- 


0) 

o 

(0 

O 


c 

o 
o 

Q. 


35 
30 
25 
20 
15 
10 

5  ^ 


1    2    3    4    5    6    7     8    9   10   11  12   13  14 

Number  of  Seeds  per  Cache 

Figure  4 — Clark's  nutcracker  cache  size  fre- 
quency. Data  from  1 57  observations  made  at 
several  locations  in  Wyoming. 


Transport  distances  varied  greatly.  Seeds  were  placed 
as  close  as  50  m  from  the  site  of  the  harvested  tree,  or 
were  transported  at  least  3.5  km  to  the  Breccia  Cliffs  on 
the  edge  of  the  Squaw  Basin  site.  With  little  seed  left  in 
the  trees  by  mid-October,  the  birds  began  retrieving  many 
of  their  caches  made  in  the  Squaw  Basin  meadows  and 
recaching  them  on  the  southwest-facing  slopes  of  the 
Breccia  Cliffs. 

Large  flocks  of  nutcrackers  would  often  cache  seed 
together.  At  Mount  Washburn,  a  flock  estimated  at 
150  birds  was  seen  caching  seed  under  an  open-grown 
whitebark  pine  stand,  with  10-15  birds  within  a  10-m^ 
area.  No  aggression  occurred  among  the  birds  during 
these  observations. 

By  November  2,  1980,  snow  covered  many  of  the  cach- 
ing areas  for  the  winter.  Nutcrackers  were  observed  on 
several  occasions  successfully  pecking  through  up  to 
25  cm  of  snow  and  ice  to  retrieve  a  cache  during  the  win- 
ter months.  The  Togwotee  Pass  area  may  receive  up  to 
1,500  cm  of  snow  a  year,  yet  the  windswept  ridges  and 
south-facing  slopes  remain  exposed  enough  for  the  birds 
to  retrieve  their  caches.  Caches  on  the  northeast-facing 
slopes  and  under  the  forest  canopy  are  more  frequently 
used  during  June,  July,  and  August,  as  the  snow  recedes 
from  these  sites  last  (Vander  Wall  and  Hutchins  1983). 

The  number  of  seed  an  individual  nutcracker  caches 
annually  has  been  estimated  in  several  studies.  The 
numbers  for  whitebark  pine  range  from  32,000  in  the 
Sierras  (Tomback  1982)  to  98,000  in  the  Rocky  Moun- 
tains (Hutchins  and  Lanner  1982).  Because  of  the  num- 
ber of  variables  that  must  be  considered  (such  as  flight 
distances  and  the  amount  of  available  seed),  these  esti- 
mates vary  a  great  deal  from  site  to  site  and  year  to  year. 

Studies  by  Tomback  (1982),  Vander  Wall  (1988),  and 
Vander  Wall  and  Balda  (1977),  estimate  that  an  indi- 
vidual bird  caches  several  times  more  seed  than  it  needs 
to  survive  through  the  winter  and  early  spring.  At  that 
time  other  food  items  become  available — although  the 
nutcrackers  continue  to  use  their  caches  heavily  until 
the  new  cone  crop  begins  to  mature  (Vander  Wall  and 
Hutchins  1983).  This  leaves  many  unused  caches  for 
potential  germination  and  establishment  of  whitebark 
pine  seedlings  (Hutchins  and  Lanner  1982). 

Clark's  nutcracker  stands  out  from  all  the  other  poten- 
tial seed  dispersers  in  two  major  ways.  First,  it  consis- 
tently disperses  seed  in  a  way  that  increases  the  chance 
of  seedling  establishment.  The  seed  is  placed  just  below 
the  soil  surface  and  hidden  from  seed  predators  (Hutchins 
and  Lanner  1982;  Lanner  1980). 

Second,  nutcrackers  scatter  caches  across  the 
landscape — both  long  and  short  distances  from  the  source 
trees.  Dispersal  distances  of  up  to  22  km  have  been  ob- 
served by  Vander  Wall  and  Balda  (1977).  Also,  the  scat- 
tering of  thousands  of  their  caches  reduces  predation. 

Steller's  Jay  (Cyanocitta  stelleri) — Having  many  of 
the  behavioral  and  physiological  attributes  of  the  Clark's 
nutcracker  (Vander  Wall  and  Balda  1981),  the  Steller's 
jay  could  have  potential  as  a  dispersal  agent  of  whitebark 
pine.  In  this  part  of  their  range,  they  were  found  to  be 
primarily  solitary  foragers  and  uncommon  visitors  to  the 
whitebark  pine  forest.  These  birds  did  not  forage  for 
whitebark  pine  seed  until  early  September  when  the 


166 


cones  dried  and  the  scales  separated  slightly.  Their  bill 
structure  does  not  approach  the  efficiency  of  the  nut- 
cracker in  prying  apart  cone  scales  to  get  to  the  seed 
(Vander  Wall  and  Balda  1981),  thus  they  were  often  only 
able  to  extract  seed  from  cones  that  had  been  exposed  by 
nutcrackers.  These  jays  harvested  whitebark  pine  seed 
from  the  ground  and  canopy  in  less  than  half  of  their  ob- 
served foraging  time  during  the  fall  of  1980  and  1981 
(Hutchins  and  Lanner  1982).  The  seed  was  always  placed 
between  their  feet  and  hammered  with  their  bill  to  crack 
the  hull;  nutcrackers  primarily  cracked  the  seed  between 
their  mandibles. 

Occasionally  Steller's  jays  pouched  their  seed  in  their 
elastic  esophagus  for  later  caching.  They  were  observed 
caching  whitebark  pine  seeds  on  seven  occasions;  how- 
ever, none  of  these  were  in  the  soil.  The  birds  placed 
the  pine  seeds  in  the  crotch  of  a  tree,  a  densely  foliated 
witches  broom,  and  under  dense  lichen  growth  along  a 
tree  branch.  The  largest  number  of  seeds  I  observed  be- 
ing pouched  at  one  time  by  this  species  was  five,  although 
data  from  Vander  Wall  and  Balda  (1981)  indicate  these 
birds  could  hold  up  to  32  seeds/pouchload.  Because  this 
species  does  not  cache  whitebark  pine  seed  in  the  soil  and 
has  a  seed  foraging  rate  comparable  to  small  passerines 
(Hutchins  and  Lanner  1982),  it  is  a  very  improbable  agent 
for  seedling  establishment. 

Raven  (Corvus  corax) — third  corvid  rarely  observed 
foraging  on  whitebark  pine  seed  was  the  common  raven. 
These  birds  had  a  great  deal  of  difficulty  extracting  seed 
from  whitebark  pine  cones  due  to  their  thick,  long  bills. 
They  dropped  most  of  the  seed  when  foraging  (Hutchins 
and  Lanner  1982).  Caching  of  carrion  was  observed  with 
this  species,  but  no  observations  were  made  of  seed  cach- 
ing. In  Asia,  Reimers  (1959  cited  in  Tiircek  and  Kelso 
1968)  described  observations  of  ravens  caching  Japanese 
{Pinus  pumila)  and  Siberian  stone  pine  (P.  sibirica),  al- 
though it  was  not  clear  in  what  type  of  substrate. 

Other  Passerines — Pine  grosbeaks  (Pinicola  enuclea- 
tor)  primarily  foraged  on  whitebark  pine  seed  that  had 
been  exposed  by  Clark's  nutcrackers  breaking  off  cone 
scales.  Their  large  conical  beaks  also  enabled  them  to 
tear  away  cone  scales  to  get  at  seed.  Their  foraging  rates 
were  very  slow  compared  to  nutcrackers  due  to  their  tech- 
nique and,  as  a  consequence,  they  had  little  influence  on 
the  depletion  of  the  whitebark  pine  cone  crop. 

Grosbeaks  are  not  known  to  cache  food  (Smith  and 
Balda  1979;  Vander  Wall  1990).  During  my  observations 
of  this  species,  they  cracked  the  seed  coat  and  consumed 
the  seed  in  the  tree  crown.  It  is  highly  unlikely  that  they 
covild  pass  an  intact  seed  through  their  digestive  tract. 

The  moxmtain  chickadee  (Parus  gambeli)  and  red- 
breasted  nuthatch  (Sitta  canadensis)  occasionally  searched 
through  the  whitebark  pine  cones  during  September  after 
the  seeds  were  exposed  by  nutcrackers.  Both  these  spe- 
cies almost  always  dropped  the  seed  from  the  tree  crown, 
as  the  seeds  were  too  large  for  them  to  handle.  Their  un- 
successful foraging  bouts  contributed  to  the  small  percent- 
age of  seed  found  on  the  soil  surface,  which  was  later  con- 
sumed by  other  animals.  Neither  bird  species  was  observed 
caching  whitebark  pine  seed,  although  they  are  known  to 


cache  smaller  seeded  conifer  species  imder  bark  (Smith 
and  Balda  1979). 

MAMMAL  DISPERSERS 

Red  Squirrel  (Tamiasciurus  hudsonicus) — Red 

squirrels  are  common  denizens  of  whitebark  pine  forests. 
In  fact  they  were  the  second  most  commonly  observed  ver- 
tebrate next  to  the  Clark's  nutcracker  (Hutchins  and  Lan- 
ner 1982).  These  mammals  actively  defend  their  territo- 
ries (Smith  1968),  and  like  nutcrackers,  harvest  whitebark 
pine  cones  and  seeds  as  a  fall  and  winter  food  source.  Red 
squirrels  did  not  occiir  on  the  meadow  sites. 

Red  squirrels  spent  most  of  their  foraging  time  (75.8  per- 
cent) on  whitebark  pine  cones  or  recovering  dropped 
whitebark  pine  seeds.  This  includes  the  time  it  takes 
to  place  the  cones  in  the  midden.  Most  of  the  rest  of  the 
observed  foraging  activities  were  collecting  Engelmann 
spruce  cones  (11.3  percent)  and  harvesting  seeds  of  herba- 
ceous plants  (12.9  percent). 

Red  squirrels  began  harvesting  cones  as  early  as  nut- 
crackers harvested  seeds  (Jvily  13,  1979),  but  did  so  more 
intensely  during  July  while  the  nutcrackers  relied  more 
heavily  on  the  previous  year's  caches  (fig.  3).  Their  forag- 
ing rate  is  much  higher  than  that  of  the  other  mammals 
discussed  in  this  paper  because  they  usually  harvest  en- 
tire cones.  Most  of  the  time,  they  remove  the  cone  by  pull- 
ing it  off  the  branch  with  their  teeth,  instead  of  cutting 
the  subtending  branch  (Hutchins  1990).  Thus,  cone  re- 
moval is  likely  to  cause  little  change  in  tree  growth  form 
or  cone  production. 

Cone  caching  began  on  August  4  at  Squaw  Basin. 
Red  squirrels  often  would  leave  cones  on  the  groimd  below 
the  tree  where  they  were  cut  for  periods  of  up  to  3  days 
before  caching  them  in  their  midden.  Caching  of  other 
conifer  cone  species  began  later  with  Engelmann  spruce 
on  August  18  in  1980;  subalpine  fir  (September  11  in 
1980),  and  lodgepole  pine  (September  27  in  1980).  All 
cones  were  stored  in  "middens"  {n  =  114).  These  middens 
can  be  quite  extensive  and  are  composed  of  many  years  of 
cone  debris  above  the  soil  surface  (Finley  1969;  Reinhart 
and  Mattson  1990;  Smith  1970).  Of  the  time  spent  cach- 
ing food,  61.4  percent  was  devoted  to  whitebark  pine 
cones  and  16.9  percent  to  whitebark  pine  seeds.  Foraging 
on  the  cones  of  other  conifers  in  the  subalpine  made  up 
most  of  the  rest  of  the  caching  time  (7,304  seconds  of  ob- 
servation). Red  squirrels  also  cached  some  mushrooms 
and  herbaceous  seeds. 

Squirrels  were  also  found  to  cache  whitebark  pine  seeds 
in  their  midden  beginning  aro\md  September  16,  1980. 
These  seed  caches  were  placed  6.5-  to  40-cm  deep  {n  =  6); 
four  of  these  observations  were  between  11  and  11.5  cm 
deep.  The  number  of  seeds  per  cache  ranged  from  14  to 
55  ix  =  28.8;  19.2  seeds/cache;  n  =  4),  although  Kendall 
(1981)  foimd  up  to  176  seeds  in  a  single  hole.  I  examined 
the  seeds  from  two  squirrel  caches  gind  found  all  the  seeds 
sound. 

Red  squirrels  actively  chased  Clark's  nutcrackers  from 
the  trees  above  their  territories  and  their  middens.  On 
two  occasions,  however,  I  watched  nutcrackers  steal  whiter 
bark  pine  seeds  from  squirrel  middens. 


167 


Chipmunk  (Eutamias  sp.) — Chipmtinks  are  uncom- 
mon visitors  to  whitebark  pine  tree  crowns  but  do  occur 
on  both  open-meadow  and  forested  sites.  Most  of  their 
time  is  spent  on  the  ground,  usually  near  cover  plants  like 
sagebrush  (Artemisia  tridentata).  They  preferred  herba- 
ceous plant  parts  (lupine  seed,  grasses,  etc.)  on  my  subal- 
pine  study  sites  until  the  third  week  of  September  when 
these  plants  died  back  and  presimiably  lost  much  of  their 
nutritive  value.  At  that  time,  they  clumsily  foraged  on 
the  remaining  whitebark  cone  crop  (-10  percent  of  the 
seed  was  left  in  the  forest,  -40  percent  in  the  meadow). 
They  more  commonly  foraged  on  the  smsdl  amount  of  seed 
below  the  trees,  which  had  been  dropped  by  any  of  the 
species  discussed  earlier.  Tevis  (1953)  and  Heller  (1971) 
similarly  foimd  that  in  the  Sierra  Nevada  chipmxmks 
devote  little  time  to  foraging  on  whitebark  pine  seed  and 
cones. 

I  found  no  evidence  of  chipmimks  caching  whitebark 
pine  seed.  Broadbooks  (1958)  found  the  larder  cache 
depth  of  yellow  pine  chipmimks  {Eutamias  amoenus)  to 
average  28  cm.  They  use  these  types  of  caches  as  a  winter 
and  spring  food  source  when  they  are  periodically  aroused 
during  hibernation  (Vander  Wall  1990).  I  examined  chip- 
munk burrows  to  a  depth  of  20  cm  without  finding  any 
evidence  of  them  harvesting  and  caching  whitebark  pine 
seed. 

Yellow  pine  chipmunks  are  known  to  scatter-hoard 
smaller  amounts  of  Jeffrey  pine  (Pinus  jeffreyi)  and  pon- 
derosa  pine  {P.  ponderosa)  seed  in  shallower  caches  from 
1  to  25  mm  in  depth  (Vander  Wall  1992a).  These  species 
of  chipmunks  in  these  studies,  along  with  other  rodents, 
may  retrieve  many  caches  in  the  fall  in  the  Sierra  Nevada. 
They  use  the  seed  to  stock  their  winter  larders,  but 
many  seeds  survive  and  germinate  the  following  spring 
(Vander  Wall  1992b).  Although  chipmunks  may  play 
a  role  in  the  afforestation  of  certain  pines,  there  are  no 
data  that  support  this  conjecture  for  whitebark  pine. 
Vander  Wall  (1992a)  provides  strong  evidence  that  Jeffrey 
pine  may  arise  from  chipmunks  (Tamias  amoenus,  specio- 
sus,  and  quadramaculatus)  scatter-hoarding  seed.  In 
Japan,  Hayashida  (1989)  found  Siberian  chipmunks 
(T.  sihiricus)  were  not  an  effective  dispersal  agent  of 
Korean  stone  pine  {P.  koraiensis)  in  a  plantation.  In  this 
study,  the  limited  amount  of  seed  these  animals  harvest 
(see  seed  crop  depletion  section  of  this  paper)  apparently 
precludes  them  from  being  of  significance  to  whitebark 
pine  establishment  in  the  Rocky  Mountains.  More  detailed 
observations  of  chipmunks  in  whitebark  pine  ecosystems 
are  needed  to  more  accurately  define  their  role  as  cache 
predator  and  possible  seed  disperser. 

Golden-Mantled  Ground  Squirrel  (Spermophilus 
lateralis) — Due  to  a  slow  foraging  rate  and  relatively 
rare  occurrence  in  tree  canopies,  the  golden-mantled 
ground  squirrel  also  consumes  a  limited  amount  of  white- 
bark pine  seed.  This  species  was  found  in  a  whitebark 
pine  tree  only  once  in  188  hours  of  observation.  It  does 
collect  a  limited  amount  of  seed  from  the  ground,  which 
falls  from  cones  during  foraging  accidents  involving  other 
animals. 

Like  chipmunks,  groimd  squirrels  are  known  to  scatter- 
hoard  shallow  seed  caches  as  well  as  make  a  deeper  lar- 
der hoard  (MacClintock  1970;  Vander  Wall  1990).  This 


species  also  begins  hibernation  quite  early  in  September 
and  would  not  have  time  to  acquire  much  seed  for  storage. 
As  with  the  chipmimk,  few  studies  have  been  performed 
to  define  ground  squirrel  foraging  and  caching  behavior. 

SECONDARY  DISPERSERS 

Nocturnal  Rodents — This  group  of  animals,  made  up 
primarily  of  mice  and  voles  (family  Cricetidae),  must  also 
be  considered  as  potential  dispersal  agents  of  whitebark 
pine  seed.  They  were  not  directly  observed  in  this  study, 
but  possible  evidence  of  their  foraging  on  whitebark  pine 
seed  was  discovered  by  the  shelled  seed  left  behind  on  my 
simulated  cache  experiments. 

Smface  seed  caches  simulating  the  seed  that  was  foimd 
on  the  ground  indicate  it  will  not  last  long  (Hutchins  1989; 
McCaughey  and  Weaver  1990).  Most  of  the  seed  in  these 
caches  was  consumed  within  2  weeks  after  placement  un- 
der trees  by  evidence  of  the  seed  shells  left  behind.  Al- 
most all  of  the  shells  were  left  behind  at  the  cache  site, 
indicating  little  if  any  caching  was  done. 

There  are  two  primary  places  where  cricetid  rodents  may 
obtain  whitebark  pine  seeds:  (a)  from  seed  that  falls  to  the 
ground  and  (b)  the  discovery  of  animal  (primarily  nut- 
cracker) caches.  A  small  amount  of  seed  (-4  percent  of 
the  seed  crop)  was  found  to  fall  to  the  forest  floor  in  the 
mast  year  of  1980.  About  69  percent  of  seed  found  on  the 
ground  was  determined  imviable  by  examination  of  the 
contents  inside  the  shell.  Although  these  rodents  may 
recache  the  seed,  the  seed  was  probably  already  placed 
in  a  suitable  site  for  seedling  establishment  by  a  Clark's 
nutcracker.  Thus,  even  if  these  cricetids  relocate  the  seed 
to  another  cache,  their  positive  effect  on  the  establish- 
ment of  whitebark  pine  is  questionable  at  best. 

Abbott  and  Quink  (1970),  working  with  the  winged 
eastern  white  pine  {Pinus  strobus)  seeds,  showed  most 
caches  by  these  rodents  were  made  less  than  15  m  from 
the  seed  source.  Thus,  the  habit  of  whitebark  pine  trees 
pioneering  open  meadows  and  disturbed  areas  is  highly 
unlikely  to  arise  from  cricetid  caches.  Their  study  also 
showed  that  of  those  caches  not  recovered  by  the  time  the 
seed  germinated,  the  germinated  seedlings  were  soon  con- 
sumed by  these  animals.  This  information  coupled  with 
the  small  amount  of  seed  available  on  the  ground  for 
these  rodents  suggests  that  they  could  rarely  be  respon- 
sible for  seedling  establishment.  Future  studies  need  to 
look  at  this  group  of  dispersal  agents  more  closely  to  fur- 
ther delineate  their  role  in  whitebark  pine  ecosystems. 

Mice  of  the  genus  Peromyscus  are  members  of  this 
group  of  rodents,  which  is  known  to  scatter-hoard  pine 
seedhngs  (Abbott  and  Quink  1970;  Vander  Wall  1992a). 
Like  the  chipmunks  and  ground  squirrels,  this  species 
searches  for  seeds  on  the  ground  and  will  either  eat  them 
or  cache  them  for  later  use.  Because  so  few  sound  white- 
bark seeds  make  it  to  the  ground,  it  appears  the  role  of 
these  small  rodents  in  whitebark  pine  establishment  is 
insignificant. 

Grizzly  and  Black  Bears — Bears  {Ursus  arctos  and 
U.  americana)  primarily  obtain  seed  from  squirrel  middens 
(Kendall  1981,  1983;  Mattson  and  Jonkel  1989),  although 
black  bears  are  known  to  also  break  branches  to  harvest 
the  seed  (Tisch  1961).  Examinations  of  bear  scat  from 


168 


both  species  have  yielded  only  a  few  intact  whitebark  pine 
seeds  (Hutchins  and  Lanner  1982).  Even  if  these  seeds 
could  germinate  in  the  bear  scats,  they  would  produce  an 
insignificant  number  of  whitebark  pine  seedlings  that  are 
dispersed  a  very  short  distance  fi'om  the  squirrel  midden 
the  bear  raided. 

PERCENT  OF  SEED  HARVESTED 

On  forested  sites,  about  63  percent  of  the  seed  was  har- 
vested by  red  squirrels  and  36  percent  by  Clark's  nut- 
cracker. The  other  1  percent  was  harvested  by  all  other 
animals  combined  (table  1). 

In  contrast,  on  the  meadow  sites,  which  have  too  few 
trees  for  squirrels  to  exist,  nutcrackers  harvested  almost 
the  entire  crop  of  whitebark  pine  (99  percent).  Details  for 
how  these  values  were  calculated  are  described  in  Hutchins 
(1982,  1990). 

In  all  but  a  year  with  a  super-abundant  cone  crop,  ani- 
mals harvest  nearly  all  of  the  seed  crop  by  early  November 
in  the  Rocky  Moimtains  (Hutchins  and  Lanner  1982; 
Vander  Wall  1988).  By  this  time,  no  cones  remain  on 
the  trees  and  seeds  do  not  have  a  chance  to  survive  long 
enough  on  the  grotmd  £ind  germinate  as  suggested  by  Day 
(1967)  and  others.  In  this  study  I  found  that  the  seed  and 
cones  that  have  been  dropped  to  the  groimd  are  eaten 
within  3  weeks  by  various  foragers. 

CONCLUSIONS 

Nutcrackers,  squirrels,  and  the  other  whitebark  pine 
seed  forager  guilds  will  occasionally  drop  some  seed  to  the 
ground  (~4  percent).  Some  seed  is  dropped  intentionally 
(Vander  Wall  and  Balda  1977)  and  some  by  accident. 
Most  of  this  seed,  however,  is  composed  of  unfilled  or 
aborted  ovules  (69.5  percent)  and  could  never  germinate. 
Consequently,  dispersal  and  establishment  of  whitebark 
pine  by  small  rodents  like  deer  mice,  ground  squirrels, 
and  chipmunks  would  be  too  rare  an  event  for  a  tree  spe- 
cies to  survive  and  prosper. 

We  also  cannot  assert  red  squirrels  to  be  an  establisher 
of  whitebark  pine  seedlings.  As  I  stated  in  earlier  work, 
I  found  no  evidence  of  whitebark  pine  establishing  on 


squirrel  middens  (Hutchins  1989).  The  midden  "seed  bed" 
is  too  deep  in  organic  debris,  and  the  activity  on  the  midden 
constantly  distvu-bs  the  soil. 

Squirrels  do  harvest  most  of  the  whitebark  pine  seed 
in  the  forest,  and  they  are  the  major  seed  predator  on  for- 
ested sites.  Their  midden  cone  stores  are  also  important 
for  providing  grizzly  and  black  bears  with  an  important 
fall  food  source  (Reinhart  and  Mattson  1990). 

Nutcrackers,  on  the  other  hand,  are  one  of  the  most  im- 
portant biotic  influences  developing  and  changing  subal- 
pine  communities.  This  species  alone  probably  accoimts 
for  nearly  all  whitebark  pine  regeneration,  except  for 
chance  happenings.  As  far  as  the  bird  and  the  tree  are 
concerned,  it  is  more  profitable  to  cache  in  the  open  mead- 
ow. Much  less  predation  occurs  on  nutcracker  caches  in 
the  meadow  (Hutchins  1990),  and  the  small  ridges  are 
usually  fi'ee  of  snow  due  to  wind  action.  Consequently, 
the  higher  cache  survival  rate  benefits  tree  regeneration 
as  well  as  the  survival  of  the  nutcracker.  This  more  than 
any  factor  may  be  why  we  see  whitebark  growing  where 
we  do — pioneering  the  exposed  ridges,  roadside  cuts, 
burned  sites,  and  meadow  swales. 

The  open-grown  meadow  trees  may  be  the  most  impor- 
tant because  the  seeds  from  these  trees  will  have  a  greater 
chance  of  being  dispersed  by  Clark's  nutcracker.  These 
small  tree  islands  lack  the  major  seed  predator  (the  red 
squirrel).  It  should  be  added  that  Reinhart  and  Mattson 
(1990)  also  foimd  lower  numbers  of  red  squirrels  using 
pure  whitebark  pine  stands.  These  also  may  be  good 
sources  for  increasing  whitebark  pine  regeneration. 

When  nutcrackers  forget  where  they  placed  a  cache  (for 
example,  Vander  Wall  1982),  or  die,  or  a  rodent  does  not 
discover  the  seed  cache — it  has  a  chance  to  germinate. 
By  placing  the  seed  in  an  excellent  germination  bed  just 
below  the  soil  surface  (2-3  cm)  and  also  hiding  the  seed 
fi'om  easy  discovery  by  seed  predators,  the  bird  may  begin 
a  new  forest  stand. 

Eventually,  these  whitebark  pine  trees  modify  the  once- 
open  subalpine  landscape  so  other  more  shade-tolerant 
species  such  as  Engelmann  spruce  and  subalpine  fir  can 
establish  themselves  in  this  commimity  (Amo  1989; 
Franklin  and  Dyrness  1973;  Snethen  1980).  The  seed 
and  cover  produced  by  whitebark  pine  attracts  a  large 


Table  1 — Estimate  of  whitebark  pine  seed  harvested  by  various  animals  in  Wyoming 


Mean  seeds 

Minutes  spent 

Foraging 

Seeds 

Number  of 

Percent  seeds  harvested 

extracted/ 

foraging/ 

days/ 

harvested/ 

individuals 

by  all  individuals 

Species 

minute^ 

day^ 

season 

individual 

visiting  trees 

Forest 

Meadowr 

Clark's  nutcracker 

7.9 

180 

91 

129,402 

448 

36.3 

99.4 

Steller's  jay 

.7 

120 

55 

4,620 

11 

<.1 

.1 

Raven 

.6 

30 

53 

954 

15 

<.1 

<.1 

Noncorvids 

.7 

120 

56 

4,704 

43 

.1 

.3 

Red  squirrel 

43.4 

240 

84 

874,944 

116 

63.5 

Chipmunk 

1.7 

120 

35 

7,140 

10 

<.1 

.1 

'Seasonal  average  from  observations  made  during  August  15  to  October  11, 1980. 
^Estimate  made  from  observed  daily  activity  patterns. 
'Meadow  area  lacks  squirrels. 


169 


number  of  vertebrates  (Amo  and  Hoff  1989;  Hutchins 
1990;  Kendall  and  Arno  1990;  Lonner  and  Pac  1990; 
Tomback  1978). 

SUMMARY 

Whitebark  pine  is  dependent  on  animal  dispersal  for 
regeneration.  Many  animals  interact  with  whitebark  pine 
seed,  but  only  Clark's  nutcracker  was  found  to  consis- 
tently disperse  seed  in  a  way  that  leads  to  the  regenera- 
tion of  this  pine.  These  long-distance  dispersal  agents 
should  be  the  central  focus  in  whitebark  pine  subalpine 
community  management. 

ACKNOWLEDGMENTS 

I  thank  Steve  Vander  Wall  for  his  insight  on  small 
mammals  and  pine  seed  dispersal.  His  comments  were 
a  great  help.  Susan  Hutchins,  Ron  Lanner,  and  Diana 
Tomback  helped  with  earlier  versions  of  this  manuscript. 
I  thank  them  for  their  efforts. 

REFERENCES 

Abbott,  H.  G.;  Quink,  T.  F.  1970.  Ecology  of  eastern  white 
pine  seed  caches  made  by  small  mammals.  Ecology. 
51:  271-278. 

Arno,  S.  1986.  Whitebark  pine  cone  crops — a  diminishing 
source  of  wildlife  food?  Western  Journal  of  Applied  For- 
estry. 1(3):  92-94. 

Arno,  S.;  Hoff,  R.  J.  1989.  Pinus  albicaulis  Engelm.: 
whitebark  pine.  In:  Burns,  R.  M;  Honkala,  B.  H.,  tech. 
coords.  Silvics  of  North  America.  Vol.  1.  Conifers.  Agric. 
Handb.  654.  Washington,  DC:  U.S.  Department  of  Agri- 
culture, Forest  Service:  269-279 

Broadbooks,  H.  E.  1958.  Life  history  and  ecology  of  the 
chipmunk  {Eutamias  amoenus)  in  eastern  Washington. 
Misc.  Publ.  103.  Ann  Arbor,  MI:  Museum  of  Zoology, 
University  of  Michigan:  5-42. 

Day,  R.  J.  1967.  Whitebark  pine  in  the  Rocky  Mountains 
of  Alberta.  Forestry  Chronicle.  43:  278-282. 

Dow,  D.  D.  1965.  The  role  of  saliva  in  food  storage  by  the 
gray  jay.  Auk.  82:  139-154. 

Finley,  R.  B.,  Jr.  1969.  Cone  caches  and  middens  ofTam- 
iasciurus  in  the  Rocky  Mountain  region.  Misc.  Publ.  51. 
In:  Contributions  in  mammalogy.  Lawrence,  KS:  Uni- 
versity of  Kansas,  Museum  of  Natural  History:  233-274. 

Franklin,  J.  F.;  Dyrness,  C.  T.  1973.  Natural  vegetation 
of  Oregon  and  Washington.  Gen.  Tech.  Rep.  PNW-8. 
Portland,  OR:  U.S.  Department  of  Agriculture,  Forest 
Service,  Pacific  Northwest  Forest  and  Range  Experi- 
ment Station.  417  p. 

Giuntoli,  M.;  Mewaldt,  L.  R.  1978.  Stomach  contents 
of  Clark's  nutcracker  in  western  Montana.  Auk. 
95:  595-598. 

Hayashida,  M.  1989.  Seed  dispersal  by  red  squirrels  and 
subsequent  establishment  of  Korean  pine.  Forest  Ecol- 
ogy and  Management.  28:  115-129. 

Heller,  H.  C.  1971.  Altitudinal  zonation  of  chipmunks 
{Eutamias):  interspecific  aggression.  Ecology. 
52(2):  312-319. 


Hutchins,  H.  E.  1982.  The  role  of  birds  and  mammals 
in  the  dispersal  and  establishment  of  whitebark  pine. 
Logan,  UT:  Utah  State  University.  67  p.  Thesis. 

Hutchins,  H.  E.  1990.  Whitebark  pine  seed  dispersal  and 
establishment:  who's  responsible?  In:  Schmidt,  W.  C; 
McDonald,  K.  J.,  comps.  Proceedings — symposiimi  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station:  245-255. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Kendall,  K.  C.  1981.  Bear  use  of  pine  nuts.  Bozeman,  MT: 
Montana  State  University.  27  p.  Thesis. 

Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  grizzly  and  black 
bears  in  the  Yellowstone  area.  Interagency  conference 
on  bear  research  and  management.  In:  Bears — their 
biology  and  management.  Calgary,  AB:  International 
Association  for  Bear  Management.  5:  166-173. 

Kendall,  K.  C;  Arno,  S.  F.  1990.  Whitebark  pine— an  im- 
portant but  endangered  wildlife  resource.  In:  Schmidt, 
W.  C;  McDonald,  K.  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  264-273. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  in  the  ecology 
and  evolution  of  limber  and  whitebark  pines.  In:  Sixth 
North  American  forest  biology  workshop  proceedings. 
Edmonton,  AB:  University  of  Alberta:  14-48. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for 
seed  dispersal  by  Clark's  nutcracker.  Canadian  Journal 
of  Forest  Research.  12(2):  391-402. 

Lanner,  R.  M.;  Vander  Wall,  S.  B.  1980.  Dispersal  of  lim- 
ber pine  seed  by  Clark's  nutcracker.  Joiirnal  of  For- 
estry. 78(10):  637-639. 

Lonner,  T.  N.;  Pac,  D.  F.  1990.  Elk  and  deer  use  of  white- 
bark pine  forests  in  southwestern  Montana:  an  ecologi- 
cal perspective.  In:  Schmidt,  W.  C;  McDonald,  K.  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
237-244. 

MacClintock,  D.  1970.  Squirrels  of  North  America.  New 
York:  Van  Nostrand  Reinhold.  184  p. 

McCaughey,  W.  W.;  Weaver,  T.  1990.  Biotic  and  micro- 
site  factors  affecting  whitebark  pine  establishment.  In: 
Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  Proceedings — 
symposivun  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989 
March  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  140-150. 

Reinhart,  D.  P.;  Mattson,  D.  1990.  Red  squirrels  in  the 
whitebark  zone.  In:  Schmidt,  W.  C;  McDonald,  K.  J., 


170 


comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
256-263. 

Smith,  C.  C.  1968.  The  adaptive  nature  of  social  organiza- 
tion in  the  genus  of  tree  squirrels  Tamiasciurus.  Eco- 
logical Monographs.  38:  31-63. 

Smith,  C.  C.  1970.  The  coevolution  of  pine  squirrels 
{Tamiasciurus)  and  conifers.  Ecological  Monographs. 
40:  349-371. 

Smith,  C.  C;  Balda,  R.  P.  1979.  Competition  among  in- 
sects, birds,  and  mammals  for  conifer  seed.  American 

Zoologist.  19:  1065-1083. 
Snethen,  K.  L.  1980.  Whitebark  pine  {Pinus  alhicaulis) 

invasion  of  a  subalpine  meadow.  Logan,  UT:  Utah  State 

University.  76  p.  Thesis. 
Tevis,  L.,  Jr.  1953.  Stomach  contents  of  chipmunks  and 

mantled  ground  squirrels  in  northeastern  California. 

Journal  of  Mammalogy.  34:  316-324. 
Tisch,  E.  L.  1961.  Seasonal  food  habits  of  the  black  bear 

in  the  Whitefish  Range  of  Northwestern  Montana. 

Missoula,  MT:  University  of  Montana.  67  p.  Thesis. 
Tomback,  Diana  F.  1978.  Foraging  strategies  of  Clark's 

nutcracker.  The  Living  Bird.  16:  123-161. 
Tomback,  Diana  F.  1982.  Dispersal  of  whitebark  pine 

seeds  by  Clark's  nutcracker:  a  mutualism  hypothesis. 

Journal  of  Animal  Ecology.  51:  451-467. 


Turcek,  F.  J.;  Kelso,  L.  1968.  Ecological  aspects  of  food 
transportation  and  storage  in  the  Corvidae.  Commimi- 
cations  in  Behavioral  Biology.  1:  277-297. 

Vander  Wall,  S.  B.  1982.  An  experimental  analysis  of 
cache  recovery  in  Clark's  nutcracker.  Animal  Behaviour. 
30:  84-94. 

Vander  Wall,  S.  B.  1988.  Foraging  by  Clark's  nutcracker 
on  a  rapidly  changing  pine  seed  source.  Condor.  90: 
621-631. 

Vander  Wall,  S.  B.  1990.  Food  hoarding  in  animals. 
Chicago,  IL:  University  of  Chicago  Press.  445  p. 

Vander  Wall,  S.  B.  1992a.  Establishment  of  Jeffrey  pine 
seedings  from  animal  caches.  Western  Jovirnal  of  Ap- 
plied Forestry.  7(1):  14-20. 

Vander  Wall,  S.  B.  1992b.  [Personal  communication]. 
July.  Reno,  NV:  University  of  Nevada,  Department 
of  Biology. 

Vander  Wall,  Stephen  B.;  Balda,  R.  P.  1977.  Coadapta- 
tions  of  Clark's  nutcracker  and  the  pinon  pine  for  effi- 
cient seed  harvest  and  dispersal.  Ecological  Monographs. 
47:  89-111. 

Vander  Wall,  Stephen  B.;  Balda  R.  P.  1981.  Ecology 
and  evolution  of  food-storage  behavior  in  conifer-seed 
caching  corvids.  Zeitschrift  fiir  Tierpsychologie.  56: 
217-242. 

Vander  Wall,  Stephen  B.;  Hutchins,  Harry  E.  1983. 
Dependence  of  Clark's  nutcracker,  Nucifraga  Colum- 
biana, on  conifer  seeds  during  the  postfledging  period. 
The  Canadian  Field  Naturahst.  97(2):  208-214. 


171 


SIZE  OF  PINE  AREAS  IN  RELATION 
TO  SEED  DISPERSAL 


Hermann  Mattes 


Abstract — Pine  seeds  are  either  wind-  or  bird-  (animal-) 
dispersed.  Only  well-adapted  seeds  support  large  ranges,  which 
are  of  similar  size  in  both  types  of  dispersal.  Seeds  adapted  to 
wind  dispersal  should  have  a  ratio  of  seed  weight  to  wing  length 
below  1.8  mg/mm.  Bird-dispersed  seeds  have  a  weight  between 
90  and  550  mg.  Species  with  seeds  above  100  mg  but  without 
zoochorous  features  have  small  or  even  relictic  ranges.  Large 
seeds  are  advantageous  under  severe  conditions  and  in  competi- 
tion. Consequently,  the  bird-pine  mutualism  contributes  to 
larger  ranges  of  those  pine  species. 


Distribution  ranges  of  plant  and  animal  species  are  de- 
termined by  four  main  factors:  ecological  potency  of  the 
species;  effectiveness  of  seed  dispersal;  size  of  area  avail- 
able; and  time  available.  Ecological  studies  on  plants  and 
vegetation  mostly  focus  on  site  conditions.  Indeed,  compe- 
tition for  nutrients,  water,  light,  and  so  on  most  often 
turn  out  to  be  the  main  factor  limiting  distribution.  How- 
ever, first  of  all,  seeds  must  be  able  to  reach  the  site  in 
question.  Therefore,  seed  dispersal  is  an  important  factor 
in  plant  ecology.  Patchy  and  irregular  patterns  of  species 
on  ruderal  or  fallow  ground  are  a  result  of  dissemination. 
Otherwise  vegetation  history  in  the  Holocene  gives  evi- 
dence that  time  since  last  glaciation  was  not  long  enough 
for  full  expansion  of  all  species.  We  expect  effectiveness 
of  seed  dispersal  to  be  correlated  with  rapidness  of  expan- 
sion and  with  size  of  range. 

Small,  winged  seeds  are  easily  disseminated  by  wind, 
which  is  present  almost  everywhere  and  emytime.  Com- 
petition of  the  seedling  within  a  closed  vegetation  cover 
as  well  as  harsh  environmental  conditions  require  large 
seeds  with  a  high  amount  of  nutrient  reserves.  Many 
pine  species  have  developed  an  almost  obligate  mutualis- 
tic  relationship  to  birds  (Nucifraga,  Gymnorhinus)  for 
seed  dispersal.  Pines  with  bird-dispersed,  large  seeds  be- 
came dominating  tree  species  at  arctic  and  alpine  timber 
lines,  in  xeric  environments  as  well  as  in  mesic  lowland 
forests. 

The  genus  Pinus  with  more  than  100  species  provides 
a  good  chance  to  test  effectiveness  of  seed  dispersal.  All 
pine  seeds  are  of  a  rather  similar  structure;  the  most  obvi- 
ous differences  are  seed  size  and  wing  length.  The  pri- 
mary type  is  very  likely  an  anemochorous  seed  with  a 
long  wing.  Second,  many  species  have  developed  large 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Hermann  Mattes  is  Professor,  Institut  fur  Geographie,  Abteilung 
Landschaftsokologie,  Westfalische  Wilhelms-Universitat,  Robert-Koch- 
Str.  26,  W-4400  MOnster,  Germany. 


seeds  with  shortened  or  missing  vnngs.  These  seeds  are 
disseminated  mainly  by  birds,  or  occasionally  by  squirrels. 
Apart  from  that,  seeds  and  cones  show  many  morphologi- 
cal and  phenological  adaptations  for  dispersal,  which  are 
discussed  elsewhere  (Lanner  1980,  1982;  Vander  Wsdl  and 
Balda  1977). 

In  this  paper  some  relations  of  seed  characteristics  and 
size  of  ranges  are  discussed.  The  main  hjrpothesis  is  to 
find  the  larger  ranges  within  pine  species  well  adapted 
to  seed  dispersal. 

MATERIALS  AND  METHODS 

From  65  out  of  about  106  pine  species  in  total  enough 
detailed  information  is  available  for  a  comparison  of  dis- 
persal abilities  and  range  sizes.  The  pine  species  are 
numbered  following  Mirov  (1967). 

Distribution  ranges  are  of  a  complex  nature.  Effective- 
ness of  dispersal  would  be  measured  best  by  the  propor- 
tion of  the  potential  area  that  really  has  been  occupied  by 
the  species.  Unfortunately,  we  do  not  know  exactly  the 
potential  ranges  of  almost  all  plant  species.  An  interest- 
ing idea  would  be  to  measure  the  distances  or  Eireas  occu- 
pied after  the  end  of  the  last  glaciation.  However,  refugials 
and  paths  of  dispersal  are  very  incompletely  known. 
What  we  can  clearly  see  are  area  and  distances  within 
ranges.  However,  it  seems  not  to  be  adequate  to  use  the 
area  of  a  range.  It  is  highly  influenced  by  size  of  the  re- 
gion (the  continent  or  vegetation  zone  that  is  inhabited 
by  the  species  concerned).  Error  should  be  less  using  the 
distance  between  the  outermost  points  of  a  range.  This 
was  calculated  as  the  orthodrome  distance. 

Split  or  disjunct  ranges  have  been  treated  in  the  same 
way  as  continuous  ranges.  It  has  proved  to  be  too  difficult 
to  decide  whether  an  interrupted  range  was  due  to  natu- 
ral factors  or  to  hxmian  influences,  especially  in  regions 
that  were  densely  populated  for  a  long  time  such  as  East 
Asia  and  the  Mediterranean. 

In  the  present  study  all  105  pine  species  listed  in  Mirov 
(1967),  and  in  addition  Pinas  longaeva,  are  considered. 
Subspecies  have  not  been  considered.  Reinge  maps  £ind 
information  was  taken  from  Kriissmann  (1968),  Little 
(1971),  Meusel  and  others  (1965),  and  Mirov  (1967). 

In  only  a  few  experiments  flying  ability  of  pine  seeds 
was  examined  (Lanner  1985;  Muller-Schneider  1977). 
To  estimate  dispersal  abilities  of  numerous  species,  indi- 
rect methods  have  been  used.  Simple,  but  reliable  meas- 
lu-ements  for  anemochorous  seeds  are  weight  of  cleaned 
seeds  and  wing  length.  Such  data  are  available  for  most 
of  the  species  (Kriissmann  1972;  USDA  1974).  The  ratio 
of  seed  weight  to  wing  length  is  used  as  an  indicator  for 
potency  of  dispersal.  Volume  and  surface  of  the  seed  and 
shape  and  surface  of  the  vring  would  be  of  great  interest, 


172 


but  these  data  are  not  available  for  most  species.  Zooch- 
orous  species  are  even  more  difficult  to  evaluate.  Seed 
and  cone  characteristics  of  species  fully  developed  for  bird 
dispersal  are  compared  morphologically  with  less-adapted 
species. 

DISPERSAL  OF  PINE  SEEDS 

Seed  dispersal  either  by  wind  or  by  birds  has  caused 
specific  adaptions.  Most  of  the  pine  species  can  be  attrib- 
uted to  these  two  main  types.  A  narrow  seed-weight  to 
wing-length  ratio  is  considered  to  be  favorable  to  flight 
ability.  From  figure  1  with  46  pine  species  Usted  we  can 
assume: 

•  In  general,  the  length  of  seed  wings  increases  until 
seed  weight  reaches  about  90  mg  (P.  palustris:  93  mg/ 
36  mm);  however,  the  ratio  of  seed  weight  to  wing  length 
increases  obviously  with  increasing  seed  weight. 

•  Winged  seeds  of  more  than  about  100  mg  have  an 
unfavorably  high  ratio  of  seed  weight  to  wing  length;  wing 
length  is  insvifficient  even  in  species  with  huge  cones  and 
long  cone  scales.  These  seeds  are  expected  to  have  no 
adequate  flight  abilities. 

•  The  smallest  seed  weight  of  a  wingless  seed  species 
{P.  flexilis)  is  93  mg;  most  of  the  species  with  heavier 
seeds  are  wingless,  or  seed  wings  separate  easily  (P.  penta- 
phylla  No.  83,  P.  himekomatsu  No.  84,  P.  strobiformis 
No.  35;  also  in  P.  pinea  No.  73). 

Comparing  morphological  features  with  area  size  we 
can  conclude  the  following  (fig.  2,  table  1). 

•  Wind-dispersed  pine  species  with  a  low  ratio  of  seed 
weight  to  wing  length  often  have  larger  distribution 
ranges  than  species  with  a  high  ratio. 

•  There  is  evidence  for  a  log-linear  relationship  for  the 
largest  ranges  of  each  class  of  seed  weight/wing  length  ra- 
tio. Species  with  a  ratio  higher  than  2  mg/mm  are  less 
distributed  on  an  average.  It  is  supposed  that  dispersal 
abilities  limit  the  range  of  distribution.  A  remarkable  ex- 
ception is  P.  pinea  (No.  73).  Its  origin  is  not  known,  and 
it  has  been  cultivated  in  many  places  since  Phoenicean 
times,  approximately  4,000  years  ago.  Therefore,  the 
whole  present  range  had  to  be  taken  into  consideration. 

•  Many  species  do  not  reach  an  area  as  large  as  could 
be  expected  in  view  of  their  theoretical  abilities  for  seed 
dispersal.  This  can  be  partly  explained  by  geographical 
and  ecological  reasons.  Species  of  southeastern  North 
America  (Nos.  22-31)  depend  on  a  relatively  small  area 
limited  by  the  ocean  or  woodless  plains.  The  range  of 
Cahfomian  pines  (Nos.  5,  6,  17-19)  of  the  Mediterranean 
type  is  limited  by  climate.  Mexican  pines  (seed  data  only 
for  P.  chihuahua  and  patula,  Nos.  40  and  61)  also  are  re- 
stricted geographically  and  ecologically  to  narrow  moxm- 
tainous  belts.  Pinus  peuce  and  P.  heldreichii  (Nos.  71  and 
75)  are  endemic  to  the  Balkan  mountains  in  southeastern 
Europe. 

•  Ranges  of  bird-dispersed  pines  are  of  similar  size  as 
those  of  wind-dispersed  pines;  differences  are  mainly  due 
to  the  very  restricted  range  of  the  bird-dispersed  P.  quad- 
rifolia  in  America  and  the  huge  range  of  wind-dispersed 
P.  sylvestris  in  Eurasia.  Some  bird-dispersed  pine  ranges 
are  among  the  largest  of  all  pines.  Concerning  size  of 


ranges,  dispersal  by  birds  (and  sqmrrels)  is  as  efficient 
as  dispersal  by  wind. 

•  Pine  ranges  in  America  are  smaller  than  those  in 
Eurasia.  This  is  an  effect  of  the  larger  land  mass  in  the 
Old  World  hemisphere.  Only  species  with  well-adapted 
dispersal  mechanisms  are  affected;  species  with  relicted 
ranges  show  no  difference  in  area  size. 

GENUS  PINUS  DISTRIBUTION 

Most  of  the  pine  species  are  holarctic.  Pines  occupy 
a  wide  variety  of  sites  of  which  the  most  extreme  are 
located  at  timber  lines  in  cold  and  arid  climates  as  well 
as  in  very  dry  or  perhvunid  regions  in  the  subtropics  and 
tropics.  Many  of  the  pine  species  are  restricted  to  specific 
sites.  Thus,  an  effective  dispersal  agent  is  needed  to 
reach  suitable  sites. 

Recent  centers  of  species  diversity  (fig.  3)  are  Mexico 
and  the  southwestern  States  of  the  U.S.A.  Secondary  cen- 
ters of  diversity  are  eastern  North  America,  eastern  Asia, 
and  the  Mediterranean  region.  These  were  refugial  areas 
during  glaciation.  Although  none  of  these  areas  is  neces- 
sarily the  origin  place  of  the  genus  Pinus,  recent  regions 
of  high  species  numbers  of  pines  are  assumed  to  be  places 
where  pines  have  existed  for  a  long  time. 

In  contrast,  there  are  large  areas  populated  only  by  a 
few  pine  species.  This  is  the  case  especially  in  the  present 
boreal  zone.  Also,  we  find  relatively  few  pine  species  in 
Central  America  south  of  Tehuantepec,  on  the  Caribbean 
Islands,  in  the  Himalayas,  and  in  southeastern  Asia. 
Probably,  these  areas  have  been  occupied  by  pines  for 
only  a  relatively  short  time. 

The  boreal  zone  was  reoccupied  by  pines  in  the  postgla- 
cial time,  and  we  know  that  some  species  are  stiU  expand- 
ing (for  example  see  Gorchakovsky  1993,  for  Pinus  sibi- 
rica).  Nevertheless,  their  ranges  belong  to  the  largest 
ranges  of  pine.  All  boreal  pine  species  are  among  the  spe- 
cies best  adapted  either  for  wind  dispersal  (figs.  1  and  2; 
Nos.  16,  20,  21,  32,  and  69)  or  bird  dispersal  (Nos.  67  and 
68).  Wind-dispersed  pines  have  excellent  seed-weight/ 
wing-length  ratios  from  0.4  to  1.1  mg/mm. 

Data  are  available  for  only  a  few  species  of  southern 
marginal  regions  (P.  wallichiana,  No.  91;  P.  merkusii, 
No.  101;  P.  khasya,  No.  104).  With  a  ratio  of  seed  weight/ 
wing  length  of  1.2  to  1.7  mg/mm,  adaptation  for  wind  dis- 
persal in  these  species  is  relatively  good. 

Species  that  were  more  widely  distributed  in  the  Ter- 
tiary or  early  Pleistocene,  and  that  now  occupy  relictic 
ranges  only,  show  very  different  ratios  of  seed  weight/ 
wing  length.  For  P.  aristata  (No.  5;  1.2  mg/mm)  and  per- 
haps P.  balfouriana  (No.  6;  2.1  mg/mm)  and  P.  longaeva 
(No.  6a)  seed  dispersal  might  not  be  a  restricting  factor. 
This  svirely  is  the  case  in  P.  torreyana  (No.  11). 

Species  with  reduced  seed  wings  have  less  chances  to 
be  dispersed  by  wind.  Some  of  them  are  on  the  way  to 
developing  zoochorous  features.  An  advanced  state  of 
zoochory  shows  in  especially  P.  strobiformis  (No.  35),  and 
somewhat  less  in  P.  pentaphylla  (No.  83)  and  P.  himeko- 
matsu (No.  84)  (Hayashida  1989);  they  have  large  seeds 
and  wings  that  easily  separate  fi-om  the  seed.  These  spe- 
cies are  already  mainly  disseminated  by  birds  and  ani- 
mals. Pinus  peuce  (No.  71)  has  a  small  seed  (40  mg)  and 


173 


length  of  seed  wing 


0.4 
0.4 

0.5  n 

0.4  ] 
0.4  D 
0.7 
0.6 
0.S 
0.6  h 
0.7] 
1.6  3 

1.13 
1.1  J 


0.7  1 
1.5  □ 


16 
5.9 

n  84 


seed  weight 

banks iana  32 
contorta  16 
sylvestris  69 
clausa  28 
serotina  30 
pa tula  61 
resinosa  21 
virginiana  27 
chihuahua  40 
pungens  31 
halepensis  77 
strobus  20 
nigra  74 

1.0  p  monticola  4 
attenuata  17 
heldreichii  75 

1.3  p  merkusii  101 

2.1  □  aristata  6 

1.3  ZJ  elliottii  23 
1.5  tzi  radiata  19 

ponderosa  14 
peuce  71 
engelmannii 
pinaster  80 
wall ichi ana 

□  palustris 

□  flexilis  2 

□  pumila  6  8 
m!  pentaphylla  83 
~i  jeffreyi  13 

:l  strobiformis  35 
I  albicaulis  1 
1  lamber tiana  3 


1.3 
5.7 


1.6  i 

1.9  ZD 


1.7 
2.6 


7.7 


13 


[j378 
56 


53 

91 
22 


181 


cembra  7  0 
edulis  8 
j  sibixica  67 

U  armandii  86 
i   coulteri  12 

.  1  monophylla  7 

1  gerardiana  92 
"    I  cembroides  36 
•  ^   I  quadrifolia  9 

:    "~1  koraiensis  82 

I  pinea  7  3 
:  .  '  I  sabiniana 

torievana] ll 


10 


[mml  30  20 


10 


0     0      100    200    300     400      500    600    700     800      900  [mgl 


Figure  1— Wing  length  (mm,  left),  weight  (mg,  right),  and  ratio  of  wing  length  to  weight  (mm/mg, 
middle)  of  seeds  of  46  pine  species.  Numbers  following  species  names  according  to  Mirov  (1967). 
For  sources  see  text.  Species  are  arranged  according  to  seed  weight. 


is  not  very  attractive  for  nutcrackers,  although  it  is  used 
for  hoarded  food  sometimes  as  is  P.  longaeva  (Lanner 
1988).  It  is  assumed  that  dispersal  by  nutcrackers  is  of 
some  relevance  in  establishing  these  two  species  at  tim- 
berline.  Ranges  of  both  species  are  small. 


Other  pine  species  such  as  P.  lambertiana,  P.  sabiniana, 
P.  torreyana,  P.  coulteri,  and  P.  jeffreyi  (Nos.  3,  and  10-13) 
have  large  seeds  but  lack  further  adaptations  for  disper- 
sal by  birds  and  animals.  Since  these  species  have  no  ef- 
fective dispersing  agent,  their  ranges  are  relatively  small. 


174 


{mg/mmj  in 
403  6-j 


^  148 
"5 

55 
5  7.4 

f 

2.7 
1.0 


5  -1 
4 
3  - 
2  H 

1 
0 


0.4  -1  -i 
0.6  -1.5 


11 


71 


•63 


7      36  1  86 
ftX'n   **    ****     •    ■  * 

92  8    70  35  2     8  2    6  7  6  8 


'  73 


84 


•  83 


'12 


•  3 

•  13 


•  Ca 


.  53 


-91 


•  78 
•22 

5     '19,  24';i-0l'77 
23  1  a    .  •    •  •14 

40.,7.95  'ZS^^ 

.28  26  ^52^  .32 
•l6 


•69 


5.5 
246 


6.0 
404 


6.5 
685 


7.0     7.5      8.0      8.5      9.0  9.5 


In 


1097  1808   2961    4815    8103  13360  [km] 
size  of  range 


Figure  2 — Size  of  ranges  in  relation  to  the  ratio  wing  length/ 
seed  weight  of  44  wind-dispersed  pines;  for  comparison, 
range  size  of  13  bird-dispersed  pine  species  on  top  of  the 
figure.  For  code  numbers  for  pine  species  see  appendix; 
for  sources  see  text. 


PiniLs  pinea,  as  mentioned  earlier,  is  an  exception  because 
it  has  been  cultivated  throughout  history. 

CONCLUSIONS 

Competition  in  seedlings  causes  a  high  selective  pres- 
sxire  for  large  seeds.  Even  among  wind-dispersed  pines 
seed  size  tends  to  enlarge  despite  a  strong  negative  effect 
on  dispersal  ability.  Obviously,  at  90  to  100  mg  seed 
weight  dispersal  success  by  wind  has  diminished  so  far 
that  other  agents  are  essential.  From  24  haploxylon  pine 
species,  19  species  have  wingless  or  almost  wingless  and 


large  seeds,  which  are  proven  to  be  or  expected  to  be  bird 
dispersed.  There  is  no  doubt  that  zoochory  has  developed 
in  parallel  in  several  sections  of  haploxylon  pines  (Lanner 
1989;  Tomback  and  Linhart  1990).  Three  diploxylon  spe- 
cies (P.  torreyana,  No.  11;  P.  sabiniana,  No.  12;  P.  pinea, 
No.  73)  have  large,  nearly  wingless  seeds  but  lack  further 
adaptations  for  zoochory. 

Range  sizes  of  bird-dispersed  pines  are  not  obviously 
different  from  those  of  wind-dispersed  pines.  Dissemina- 
tion of  pine  seeds  by  birds  is  at  least  as  effective  as  wind 
dispersal.  The  interaction  of  pines  and  nutcrackers  or 
jays  is  well  balanced.  Even  scattered  stands  of  zoochorous 


Table  1 — Average  size  of  ranges  (km  distance)  of  pine  species  with  different  dispersal  agents.  Wind-dispersed 
pines  are  arranged  according  to  their  seed  characteristics  (seed  weight/wing  length  in  mg/mm) 


 Distribution  ranges  

Pine  species  dispersed  by:  America  Eurasia  Global 


Birds 

1,557 

in 

=  7) 

2,680 

in  = 

6) 

2,075 

in 

=  13) 

Wind  (0.4-0.8  mg/mm) 

1,913 

in 

=  13) 

M,247 

in  = 

3) 

2,351 

in 

=  16) 

Wind  (0.9-2.0  mg/mm) 

1,926 

in 

=  9) 

2,610 

in  = 

7) 

2,225 

in 

=  16) 

Wind  (>2.0  mg/mm) 

1,121 

in 

=  7) 

21,055 

in  = 

4) 

21,097 

in 

=  11) 

Species  considered  above 

1,693 

in 

=  36) 

22,566 

in  = 

20) 

21,994 

in 

=  57) 

All  species 

1,413 

in 

=  67) 

1,779 

in= 

39) 

1,548 

in 

=  106) 

'Without  P.  sylvestris,  average  range  size  is  2,150  km. 
^Numbers  without  P.  pinea. 


175 


176 


pine  trees  support  all  needs  for  territories  of  nutcrackers 
concerning  breeding,  feeding,  overwintering,  and  social 
life.  Nutcrackers  and  jays  have  still  preserved  character- 
istic euryoecious  features  of  corvids,  especially  omnivory. 
In  Eurasia,  where  there  is  almost  no  other  animal  com- 
peting with  the  nutcracker  for  seed  hoarding,  the  range 
of  the  nutcracker  surpasses  the  ranges  of  zoochorous 
pines  in  many  regions. 

Within  wind-dispersed  pines  we  find  that  large-sized 
ranges  suppose  small  seeds  with  a  ratio  of  weight  to  wing 
length  below  approximately  1.8  mg/mm,  but  not  vice- 
versa  because  of  geographical  barriers  or  specialized  fea- 
tures of  a  species.  However,  pine  species  with  a  very  high 
ratios  of  seed  weight/wing  length  (above  1.8  mg/mm)  al- 
ways have  restricted  areas. 

Altogether,  pine  seed  dispersal  by  birds  was  very  suc- 
cessful dvuing  evolution.  Pine  species  became  able  to  oc- 
cupy additional  and  more  different  sites,  as  well  as  under 
extreme  conditions  at  the  borders  of  tree  growth  as  under 
mesic  conditions.  Consequently,  bird-dispersed  pines 
were  able  to  enlarge  their  ranges. 

REFERENCES 

Critchfield,  William  B.;  Little,  Elbert  L.,  Jr.  1966.  Geo- 
graphic distribution  of  the  pines  of  the  world.  Misc. 
Pub.  991.  Washington,  DC:  U.S.  Department  of  Agricul- 
ture, Forest  Service.  97  p. 

Gorchakovsky,  P.  L.  1993.  Distribution  and  ecology  of  Si- 
berian stone  pine  (Pinus  sibirica)  in  the  Urals.  (These 
proceedings). 

Hayashida,  Mitsuhiro.  1989.  Seed  dispersal  and  regenera- 
tion patterns  of  Pinus  parviflora  var.  pentaphylla  on 
Mt.  Apoi  in  Hokkaido.  Research  Bulletin  of  the  College 
Experiment  Forests,  Faculty  of  Agriculture,  Hokkaido 
University.  46(1):  177-190. 

Kriissmann,  Gerd.  1968.  Die  Baume  Europas.  Berlin  and 
Hamburg:  Parey  Verlag.  140  p. 

Kriissmann,  Gerd.  1972.  Handbuch  der  Nadelgeholze. 
Hambiu-g  und  Berlin:  Parey- Verlag. 

Lanner,  Ronald  M.  1980.  Avian  seed  dispersal  as  a  factor 
in  the  ecology  and  evolution  of  limber  and  whitebark 


pines.  Proc.  6th  North  American  Forest  Biology  Work- 
shop, University  of  Alberta,  Edmonton,  Canada:  15-48. 

Lanner,  Ronald  M.  1982.  Adaptations  of  whitebark  pine 
for  seed  dispersal  by  Clark's  nutcracker.  Canadian 
Journal  of  Forest  Research.  12(2):  391-402. 

Lanner,  Ronald  M.  1985.  Effectiveness  of  the  seed  wing  of 
Pinus  flexilis  in  wind  dispersal.  Great  Basin  Naturalist. 
45(2):  318-320. 

Lanner,  Ronald.  1988.  Dependence  of  Great  Basin  bristle- 
cone  pine  on  Clark's  nutcracker  for  regeneration  at  high 
elevation.  Arctic  and  Alpine  Research.  20(3):  358-362. 

Lanner,  Ronald  M.  1989.  Biology,  taxonomy,  evolution, 
and  geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposium  on  vv^hitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
14-24. 

Little,  Elbert  L.,  Jr.  1971.  Atlas  of  United  States  trees. 

Vol.  1:  Conifers  and  important  hardwoods.  Misc.  Publ. 

1146.  Washington,  DC:  U.S.  Department  of  Agriculture. 

9  p.  and  200  maps. 
Meusel,  H.;  Weinert,  E.;  Jager,  E.  1965.  Vergleichende 

Chorologie  der  zentraleuropaischen  Flora.  Bd.  I.  Jena: 

Fischer- Verlag.  9  p.  and  200  maps. 
Mirov,  Nicholas  T.  1967.  The  genus  Pinus.  New  York: 

Ronald  Press.  602  p. 
Miiller-Schneider,  Paul.  1977.  Verbreitungsbiologie 

(Diasporologie)  der  Blutenpflanzen.  Veroffentlichungen 

des  geobotanischen  Institutes  der  ETH  Zurich,  Stiftung 

Riibel.  Vol.  61.  226  p. 
Tomback,  Diana  F.;  Linhart,  Yan  B.  1990.  The  evolution 

of  bird-dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 
U.S.  Department  of  Agriculture,  Forest  Service.  1974. 

Seeds  of  woody  plants  in  the  United  States.  Agric. 

Handb.  450.  Washington,  DC.  838  p. 
Vander  Wall,  Stephen  B.;  Balda,  Russ  P.  1977.  Coadapta- 

tions  of  the  Clark's  nutcracker  and  the  pinyon  pine  for 

efficient  harvest  and  dispersal.  Ecological  Monographs. 

Durham.  47(1):  27-37. 


177 


APPENDIX:  CODING  LIST  OF  NUMBERS  OF  PINE  SPECIES  CONSIDERED 
(AFTER  MIROV  1967)  wi^c»ii^r.rtr.u 


Western  North  America 

Haploxylon  pines 

1  albicaulis 

2  flexilis 

3  lambertiana 

4  monticola 

5  balfouriana 

6  aristata 
6a  longaeva 

7  monopylla 

8  edulis 

9  quadrifolia 

Diploxylon  pines 

10  sabiniana 

11  torreyana 

12  coulteri 

13  jeffreyi 

14  ponderosa 

16  contorta 

17  attenuata 

18  muricata 

19  radiata 

Eastern  North  America 

Haploxylon  pines 

20  strobus 

Diploxylon  pines 

21  resinosa 

22  palustris 

23  elliotti 

24  toeda 

25  echinata 

26  glabra 

27  virginiana 

28  clausa 

29  rigida 

30  serotina 

31  pungens 

32  banksiana 


Mexico 

Haploxylon  pines 

35  strobiformis 

36  cembroides 

Diploxylon  pines 
40  chihuahua 
53  engelmanni 

Northern  Eurasia 

Haploxylon  pines 

67  sibirica 

68  pumila 

Diploxylon  pines 

69  sylvestris 

Mediterranean  Region 

Haploxylon  pines 

70  cembra 

71  peuce 

Diploxylon  pines 

73  pinea 

74  wi^m 

75  heldreichii 

76  montana  i=mugo) 

77  halepensis 

78  brutia 
80  pinaster 

Eastern  Asia 

Haploxylon  pines 

82  koraiensis 

83  pentaphylla 

84  himekumatsu 
86  armandi 

91  wallichiana  (=griffithi) 

92  gerardiana 

Diploxylon  pines 
95  densiflora 
101  merkusii 
104  khasya 


178 


THE  REGENERATION  PROCESS 
OF  WHITEBARK  PINE 

Ward  W.  McCaughey 


Abstract — ^Whitebark  pine  (Pinus  albicaulis  Engelm.)  regener- 
ates similarly  to  European  and  Asian  stone  pines.  Our  knowl- 
edge of  this  process  is  essential  because  whitebark  pine  is  impor- 
tant to  the  North  American  grizzly  bear  {Ursus  arctos  horribilis), 
other  wildlife  species,  hydrology  of  high-elevation  ecosystems, 
and  esthetics.  This  paper  simimarizes  available  information  on 
the  whitebark  pine  regeneration  process,  beginning  from  bud  ini- 
tiation through  germination  and  seedling  survival.  Major  factors 
limiting  germination  and  regeneration  success  are  discussed.  We 
continue  to  gain  knowledge  about  the  regeneration  process  of 
whitebark  pine,  but  further  research  is  needed  to  fully  imder- 
stand  delayed  germination  mechanisms  and  habitat  require- 
ments for  optimimi  regeneration  success. 


Whitebark  pine  (Pinus  albicaulis  Engelm.)  is  an  impor- 
tant food  source  for  the  endangered  grizzly  bear  (Ursus 
arctos  horribilis),  red  squirrels  (Tamiasciurus  hudsonicus), 
the  Clark's  nutcracker  (Nucifraga  columbiana  Wilson), 
and  a  multitude  of  other  birds  and  mammals  (Hutchins 
and  Lanner  1982;  Kendall  1983;  Knight  and  others  1987). 
It  is  also  significant  as  a  hydrologic  stabilization  plant 
and  an  esthetic  feature  of  high-elevation  communities 
(Schmidt  and  McDonald  1990).  It  provides  important 
cover  for  wildlife  and  is  used  as  an  ornamental  for  land- 
scaping but  has  only  minor  significance  as  a  timber  pro- 
ducing species  (McCaughey  and  Schmidt  1990). 

General  information  is  available  and  is  summarized 
in  this  paper  on  the  natiu-al  regeneration  process  of  white- 
bark pine,  but  more  specific  facts  are  needed  for  efficient 
management  of  this  North  American  stone  pine  (Amo  and 
Hoff  1989;  McCaughey  and  Schmidt  1990).  This  paper 
describes  the  regeneration  process  beginning  firom  bud 
initiation  throiigh  germination  and  seedling  survival. 

CONE  AND  SEED  DEVELOPMENT 

Whitebark  pine  cone  and  seed  development  begins  vnth 
cone  initiation  and  ends  with  seed  matiiration.  Climatic 
conditions  influence  the  timing  of  development  and  matu- 
ration of  cone  initiation  and  seed  maturation. 

Cone  Initiation 

The  process  of  cone  initiation  and  development  of  stami- 
nate  and  ovulate  strobili  for  whitebark  pine  has  not  been 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Ward  W.  McCaughey  is  a  Research  Forester,  Intermountain  Research 
Station,  Forest  Service,  U.S.  Department  of  Agriculture,  Bozeman,  MT 
59717-0278. 


specifically  studied.  Mirov  (1967)  and  Krugman  and 
Jenkinson  (1974)  provided  generalized  descriptions  of  the 
initiation  process  for  the  genus  Pinus.  Female  and  male 
cone  initiation  of  whitebark  pine  probably  occiu-  during 
or  just  prior  to  winter  bud  formation  from  mid-August  to 
mid-September  (Schmidt  and  Lotan  1980). 

Cone  Development 

Female  and  male  buds  overwinter  and  begin  further 
development  and  growth  in  April  and  May  depending  on 
elevation.  In  1992,  staminate  cones  matured  and  shed 
pollen  during  Jxme  and  early  Jvdy  at  high  elevations 
(2,500-2,600  m)  in  southwestern  Montana  (McCaughey 
1992)  and  in  May  and  June  at  lower  elevations  (1,550- 
1,650  m)  in  northern  Idaho  (Hoff  1992).  Mature  male 
cones  are  about  1  cm  wide  by  1  cm  long  and  female  cones 
are  1  cm  wide  by  2  cm  long  during  early  development 
when  they  are  pollen  receptive.  After  pollination  female 
cones  grow  to  a  first-year  size  of  about  3  cm  long  by  2  cm 
wide.  Female  cones  remain  on  the  tree,  while  male  cones 
fall  off  after  pollen  dispersal.  Pollination  occurs  from  late 
Jime  to  late  Jvdy,  and  fertilization  occxirs  about  12  to  13 
months  later,  simultaneous  with  rapid  cone  enlargement 
during  June  and  July  (Krugman  and  Jenkinson  1974). 

Seed  Maturation 

Following  fertilization  the  embryo  develops  and  dif- 
ferentiates into  cotyledons,  plumule,  and  radicle  (Mirov 
1967).  Whitebark  seeds  reach  maturity  between  mid- 
August  and  mid-September  depending  on  elevation  and 
climatic  conditions  during  the  ripening  process  (Krugman 
and  Jenkinson  1974;  McCaughey,  in  press  a).  The  entire 
process  from  cone  initiation  to  cone  and  seed  matiuity 
takes  about  24  months. 

Cone  and  Seed  Insects  and  Diseases 

Whitebark  pine  cones  and  seeds  are  exposed  to  insects 
and  disease  that  reduce  cone  and  seed  siirvival  during  the 
2  years  of  development  (Edwards  1990).  Cone  and  seed 
insects  that  damage  whitebark  pine  are  cone  worms  (Dio- 
ryctria  spp.  and  Eucosma  spp.),  cone  beetles  (Conophtho- 
rus  spp.)  (Bartos  and  Gibson  1990),  and  midges  and  seed 
chalcids  (Megastigmus  spp.)  (Dewey  1989).  Siroccocus 
strobilinus  Preuss  is  a  seed-borne  disease  that  kills  white- 
bark pine  seedlings  in  niu-series  and  in  natxu-al  stands. 
Calocypha  fulgens  (Pers.)  Boud.  (anamorph  =  Geniculcden- 
dron  pyriforme  Salt),  referred  to  as  a  seed  or  cold  fungus, 
may  cause  preemergence  seed  loss  (Hoff  and  Hagle  1990). 
Pre-  and  postemergence  damping-off  diseases  (Fusarium 
spp.)  may  cause  extensive  mortality,  especially  in  slowly 
emerging  seedlings  (Landis  and  others  1990). 


179 


SEED  DISPERSAL 

Whitebark  pine  cones  are  well  suited  to  the  Clark's  nut- 
cracker, a  forest  bird  that  is  the  major  disperser  of  white- 
bark  pine  seeds  (Arno  and  Hoff  1989;  Hutchins  and 
Lanner  1982;  Lanner  1980;  Lanner  and  Vander  Wall  1980; 
McCaughey  and  Schmidt  1990;  Tomback  1982). 

Cone  Attributes 

Unlike  other  conifers  such  as  western  white  pine  (Pinus 
monticola  Dougl.)  and  limber  pine  {Pinus  flexilis  James), 
whose  cone  scales  fully  open  to  release  seed,  whitebark 
pine  cone  scales  only  open  partly.  The  scale  base  does  not 
break  away  from  the  cone  axis;  the  wingless  seeds  are 
held  firmly  in  place  yet  fully  exposed  and  easily  accessible 
to  nutcracker  extraction  (Eggers  1986;  Hutchins  and 
Lanner  1982).  Nutcrackers  have  strong,  pointed  bills  like 
other  species  of  the  family  Corvidae.  Upper  portions  of 
the  cone  scales  are  easily  broken  by  nutcrackers  along 
a  thin  fracture  zone  where  the  massive  apophysis  tapers 
to  a  thin  cross  section  beneath  the  seed-bearing  cavities 
(Lanner  1982).  A  nutcracker  can  store  over  100  white- 
bark pine  seeds  in  its  sublingual  pouch,  a  saclike  modifi- 
cation on  the  floor  of  the  mouth  (Bock  and  others  1973). 

Nutcracker  Caching 

Whitebark  pine  regeneration  depends  almost  entirely 
on  nutcracker  seed  selection  and  caching  habits.  A  nut- 
cracker appears  to  discriminate  between  good,  aborted, 
insect-infected,  or  diseased  seeds  by  rattling  each  seed 
in  its  bill  before  depositing  it  in  the  sublingual  pouch 
(Hutchins  and  Lanner  1982;  Tomback  1978;  Vander  Wall 
and  Balda  1977).  Whitebark  pine  seeds  are  dispersed  by 
nutcrackers  up  to  22  km  from  their  source.  Caches  are 
buried  2  to  4  cm  deep  with  1  to  25  or  more  seeds  per  cache 
(fig.  1)  (Tomback  1978;  Vander  Wall  and  Balda  1977). 
One  nutcracker  can  store  an  estimated  22,000  to  98,000 
whitebark  pine  seeds  each  year  when  seed  is  available 
(Hutchins  and  Lanner  1982;  Tomback  1978;  Vander  Wall 
and  Balda  1977).  Estimates  of  food  requirements  indicate 
that  nutcrackers  may  store  three  to  five  times  as  many 
seeds  as  needed  (Tomback  1983). 

Secondary  Dispersers 

Nutcrackers  compete  with  other  seed  consumers  for 
whitebark  pine  seeds.  Red  squirrels  harvest  cones  and 
seeds  from  mid-July  to  early  November,  greatly  reducing 
the  availability  of  seed  for  nutcrackers  (Eggers  1986; 
Hutchins  and  Lanner  1982;  Reinhart  and  Mattson  1990; 
Smith  1968).  Squirrels  store  cones  in  middens  and  seeds 
in  caches  on  the  forest  floor  (Hutchins  and  Lanner  1982; 
Reinhart  and  Mattson  1990). 

Grizzlies  obtain  whitebark  pine  seeds  primarily  from 
squirrel  middens  (Kendall  1983).  Germination  probabil- 
ity from  squirrel-cached  cones  and  seed  is  low  due  to  their 
deep  caching  habits  (>7  cm)  and  small  numbers  of  midden 
sites.  H5^ocotyl  growth  of  whitebark  pine  is  only  3  to  4  cm; 
emergence  of  cotyledons  above  the  soil  surface  in  a 


Figure  1 — Nutcracker  cache  site  with  five  white- 
bark pine  germinants  growing  on  a  litter  seedbed 
near  a  log. 


squirrel  midden  is  unlikely  (Lanner  1982).  Squirrels 
feed  on  the  cones  and  seed  during  the  winter  and  spring 
months  thus  constantly  disturbing  the  middens  (Lanner 
1982). 

Many  other  animals  feed  on  whitebark  pine  seed  and  are 
considered  secondary  dispersers  because  of  the  low  prob- 
ability of  germination  following  their  feeding  and  caching 
(Hutchins  and  Lanner  1982;  McCaughey  and  Schmidt 
1990).  Other  common  animals  that  harvest  whitebark 
pine  seed  either  from  the  cones  directly  or  indirectly  from 
the  ground  or  other  animal  caches  are:  birds — ^William's 
sapsucker  (Sphyrapicus  thyroideus),  hairy  woodpecker 
{Picoides  villosus),  white-headed  woodpecker  {P.  albolar- 
vatus),  mountain  chickadee  (Parus  gambeli),  white-breasted 
nuthatch  {Sitta  carolinensis),  Cassin's  finch  {Carpodacus 
cassinii),  red  crossbill  {Loxia  curvirostra),  pine  grosbeak 
(Pinicola  enucleator),  Steller's  jay  {Cyanocitta  stelleri), 
raven  (Corvus  corax),  and  red-breasted  nuthatch  {Sitta 
canadensis).  Mammals — chipmimks  {Eutamias  spp.), 
deer  mouse  {Peromyscus  maniculatus),  golden-mantled 
groimd  squirrel  {Spermophilus  lateralis),  southern  red- 
backed  vole  {Clethrionomys  gapperi),  chickaree  {Tamia- 
sciurus  douglasi)  (Hutchins  and  Lanner  1982;  McCaughey 
and  Schmidt  1990),  and  black  bears  {Ursus  americanus) 
(Craighead  and  others  1982;  Kendall  1983).  Secondary 
dispersers  not  only  reduce  the  seed  crop  but  limit  the 
availability  of  seed  to  nutcrackers  and  thus  the  probabil- 
ity of  regeneration. 

Usually  whitebark  pine  cone  crops  are  depleted  by 
cone  and  seed  consumers.  Hutchins  and  Lanner  (1982) 
described  a  3-year  period  when  no  whitebark  pine  cones 
fell  to  the  ground  other  than  by  animal  clipping.  In  con- 
trast, so  many  cones  were  produced  in  1989  that  many 
were  not  utilized  by  seed  consumers  and  fell  to  the  groimd 
(McCaughey  and  others  1990).  Grizzly  bears  were  ob- 
served foraging  on  these  cones  in  1990  (Reinhart  1990). 
The  1989  mast  year  may  have  reached  the  upper  limit  of 


180 


cone  production,  a  level  that  rarely  occiirs.  Wind  storms 
may  have  dislodged  the  whitebark  cones  making  them 
available  to  ground  foraging  animals. 

The  probability  of  germination  is  low  from  seeds  that 
have  fallen  to  the  ground  in  cones  or  been  dropped  by  ani- 
mals. All  whitebark  pine  seeds  sown  on  the  ground  are 
eaten  or  taken  by  seed  consxmiers  when  not  protected 
with  exclosures  (McCaughey  1990).  This  situation  oc- 
curred in  two  successive  seasons  when  seeds  were  sown 
during  field  germination  tests.  Eight  species  of  rodents 
were  trapped  on  site  dining  the  field  germination  test. 
The  two  most  abundant  species  were  the  deer  mouse  and 
the  southern  red-backed  vole  representing  54  and  23  per- 
cent of  the  total  popidation  trapped  (McCaughey  1990). 

GERMINATION 

Information  on  seed  storage  and  laboratory  germina- 
tion of  whitebark  pine  seeds,  under  a  variety  of  seed  treat- 
ment methods,  has  been  summarized  and  discussed  by 
several  authors  in  the  symposium  proceedings  "Whitebark 
Pine  Ecosystems:  Ecology  and  Management  of  a  High- 
Mountain  Resource"  (Schmidt  and  McDonald  1990).  The 
following  sections  extract  pertinent  information  from 
those  various  soiirces. 

Seed  Storage 

Whitebark  pine  seed  has  been  frozen  imder  environ- 
mentally controlled  conditions  for  up  to  20  years  (Schubert 
1954).  Seed  viability  decreased  over  that  time  period  from 
50  percent  to  3  percent.  Mirov  (1946)  found  that  viability 
of  whitebark  pine  seed  dropped  from  24  percent  at  time 
of  collection  to  17  percent  after  being  frozen  for  8  years, 
but  dropped  to  1  percent  after  11  years  of  storage.  Viabil- 
ity of  whitebark  pine  seed,  as  related  to  storage  time,  is 
dependent  on  seed  maturity  when  harvested,  seed  han- 
dling prior  to  storage,  and  methods  and  length  of  time 
of  stratification  for  germination  tests  (McCaughey  and 
Schmidt  1990). 

Controlled  Germination 

ViabiHty  of  whitebark  pine  seed,  even  iinder  controlled 
conditions,  is  highly  variable,  ranging  from  0  to  75  per- 
cent. One  procedure  for  germinating  whitebark  pine 
seeds  is  to  soak  seeds  in  running  tap  water  for  1  to  2  days 
and  stratify  moist  at  1  to  5  °C  for  90  to  120  days  in  plastic 
bags  (Krugman  and  Jenkinson  1974).  Stratification  time 
may  vary  by  seed  maturity.  Jacobs  and  Weaver  (1990a) 
found  that  1  month  of  stratification  was  sufficient  to  in- 
crease germination  from  5  percent  to  about  40  percent; 
longer  stratification  periods  (to  8  months)  did  not  improve 
germination.  The  Coevir  d'Alene  Nursery  in  Idaho, 
U.S.A.,  used  the  following  procedm-es  for  germination 
tests  for  western  white  pine  and  whitebark  pine: 

A.  Place  seed  in  nylon  mesh  bags. 

B.  Soak  seed  for  48  hours  in  ninning  tap  water.  Place 
nylon  mesh  bag  in  plastic  bag. 

C.  Stratify  for  100  days  at  1  to  2  °C.  Within  that  100- 
day  stratification  time  resoak  the  seed  for  1  hour  each 
week. 


D.  After  100  days,  remove  seed  from  stratification  and 
surface  dry. 

E.  Using  a  vacuum  seeder,  place  100  seeds  on  moist 
paper  towels  (Kimpak)  in  each  of  four  plastic  trays. 

F.  Place  trays  in  germinator  and  take  counts. 

Seedling  Morphology 

Whitebark  pine  germinants  have  thicker  stems  than 
those  of  associated  conifers.  Stem  diameters  range  from 
2  to  4  mm  for  whitebark  and  only  1  to  2  mm  for  lodgepole 
pine  {Pinus  contorta  var.  latifolia),  Engelmann  spruce 
{Picea  engelmannii  Parry  ex  Engelm.),  and  subalpine  fir 
iAhies  lasiocarpa  [Hook.]  Nutt.)  (McCaughey  1988).  Ger- 
minants produce  five  to  12  cotyledons  and  grow  to  heights 
of  3  to  5  cm  in  the  first  growing  season  (fig.  2)  (McCaughey 
and  Schmidt  1990).  Needles  vary  in  length  from  7  to  10 
cm  with  stomata  positioned  on  the  dorsal  and  ventral 
sides.  The  hypoderm  is  weak,  usually  one  cell  in  width; 
the  endodermis  has  an  outer  wall  that  is  strongly  thick- 
ened, and  there  is  only  one  fibrovascular  bundle  (Harlow 
1931). 

Whitebark  pine  trees  commonly  have  multiple  stems 
from  forking  of  a  single  stem  or  the  merging  of  several 
seedlings  from  a  nutcracker  cache.  Forking  occurs  on  a 
small  percentage  of  whitebark  pine  germinants.  Nearly 
7  percent  of  first-year  germinants  have  two  forks  and 
17  percent  have  three  or  more  forks  (McCaughey  1988; 
Weaver  and  Jacobs  1990).  Genotypic  analyses  of  multi- 
stemmed  clumps  show  that  58,  70,  and  83  percent,  of 
three  stands  sampled  in  Alberta,  Canada,  and  Wyoming, 
U.S.A.,  had  stems  of  mixed  genetic  origin  while  42,  30, 


Figure  2 — Single  whitebark  pine  seedling  germin- 
ating on  a  mineral  seedbed  and  showing  newly 
emerged  cotyledons. 


181 


and  17  had  arisen  by  branching  (Furnier  and  others 
1987;  Jacobs  and  Weaver  1990b;  Linhart  and  Tomback 
1985).  This  indicates  that  trees  of  mixed  genetic  origin 
had  merged  from  several  genetically  individual  seedlings 
from  the  same  cache. 

Whitebark  pine  germination  and  root  growth  are  affected 
by  early  season  soil  temperatures.  Seeds  germinate  when 
soil  temperatures  are  between  10  and  40  °C  with  an  opti- 
mum temperature  range  of  25  to  35  °C  (Jacobs  1989; 
Jacobs  and  Weaver  1990a).  Root  growth  occurs  between 
10  and  45  °C  with  the  optimum  temperature  range  being 
25  to  35  °C.  This  optimum  range  provides  conditions 
where  whitebark  roots  grow  5  to  15  mm  per  day  (Jacobs 
and  Weaver  1990a).  First-year  root  growth  ranges  be- 
tween 5  and  18  cm  for  nutcracker-cached  seedlings  grow- 
ing in  a  forest  environment  (McCaughey  1988). 

Period  of  Germination 

Germination  from  a  single  sowing  or  caching  of  white- 
bark pine  seeds  can  continue  throughout  the  growing  sea- 
son and  during  the  next  2  to  3  or  more  years  (McCaughey, 
in  press  b).  Because  nutcrackers  bury  seed  2  to  4  cm 
deep,  seeds  may  germinate  but  not  emerge  above  the  soil 
surface  for  several  days  after  germination  or  they  may  not 
emerge  at  all.  The  term  germinant  in  this  paper  refers 
to  an  emergent,  since  it  is  difficult  to  discern  if  a  seed  ger- 
minated yet  died  prior  to  emergence. 

Germination  Sites 

Whitebark  pine  germinates  on  a  variety  of  seedbeds, 
almost  entirely  dependent  on  the  choice  of  the  nutcracker. 
Whitebark  pine  grows  within  38  habitat-phase  combina- 
tions in  eastern  Idaho  and  western  Wyoming  (Steele  and 
others  1983)  and  44  in  Montana  (Pfister  and  others  1977). 
Typical  sites  for  Clark's  nutcracker  caches  are  well-drained 
and  moist  substrates,  bare  soil,  forest  litter,  gravel,  rubble, 
in  small  plants  and  logs,  in  cracks  and  fissures  on  exposed 
rock,  and  in  pumice  soils  (Hutchins  and  Lanner  1982; 
Tomback  1978,  1982;  Tomback  and  others  1990).  Regen- 
eration is  commonly  found  on  burned  litter  seedbeds  fol- 
lowing natural  and  prescribed  burns  (Morgan  and  Bimting 
1990;  Tomback  1986), 

Nutcrackers  cache  whitebark  seeds  on  all  aspects,  but 
the  majority  of  cache  sites  occur  on  southeast,  south, 
southwest,  and  west-southwest  aspects  (Tomback  and 
others  1990).  Seed  storage  in  south-aspect  windblown 
sites  probably  ensures  that  some  caches  are  snowfree  in 
winter  and  spring  so  the  nutcracker  can  retrieve  them. 
Even  though  fewer  seeds  are  cached  on  the  north  aspects, 
regeneration  densities  are  highest  on  north  aspects.  This 
may  be  due  to  less  successful  retrieval  by  the  nutcracker, 
or  it  may  be  a  result  of  favorable  environmental  condi- 
tions such  as  moisture  and  insolation  protection  during 
the  germination  period  (Tomback  and  others  1990).  The 
tallest  and  best  formed  whitebark  pine  trees  are  often 
found  in  high  basins  or  on  gentle  north  slopes  (Arno  and 


Hoff  1989).  Drought  and  insolation  mortality  of  seedlings 
on  south  slopes  may  contribute  to  the  disproportionate 
success  of  whitebark  regeneration  on  north  aspects. 

First-Year  Germination 

Whitebark  pine  seed  germinating  from  caches  by  the 
nutcracker  begins  about  mid-June  and  continues  through 
early  September.  The  number  of  whitebark  pine  seeds 
germinating  in  the  first  year  after  caching  varies  by  year, 
probably  in  response  to  timing  of  spring  snowmelt  and 
early  summer  rains  (fig.  3)  (McCaughey  1990).  For  ex- 
ample, in  1988  only  11.5  percent  of  buried  seeds  germi- 
nated, probably  due  to  below-average  precipitation  (82 
percent  of  normal).  In  1989  precipitation  was  near  normal 
and  germination  of  1988  sown  seed  was  nearly  34  percent 
(table  1)  (McCaughey  1990). 

First-year  germination  of  whitebark  pine  is  significant- 
ly affected  by  percent  shade  cover  (0,  25,  and  50  percent), 
seedbed  condition  (mineral,  litter,  and  burned),  and  sow- 
ing depth  (surface  sown  and  buried)  (table  1)  in  controlled 
field  studies  in  southwestern  Montana,  U.S.A.  (McCaughey 
1990).  Percent  germination  of  whitebark  pine  was  signifi- 
cantly ip  =  0.008)  higher  (20.4  percent)  under  a  50  percent 
shade  cover  than  with  no  shade  (16.7  percent)  (table  1) 
(McCaughey  1990).  Germination  of  whitebark  pine  seeds 
sown  in  1988  was  26  percent  on  mineral  seedbeds  and  sig- 
nificantly less  on  litter  and  burned  seedbeds,  14.5  and 
15.3  percent,  respectively  (McCaughey  1990).  Buried 
whitebark  pine  seeds  (2  to  4  cm  deep)  had  significantly 
higher  germination  in  1988  and  1989  (11.5  and  33.7  per- 
cent, respectively)  than  surface-sown  seeds  that  were  pro- 
tected from  seed  consumers  in  the  same  years  (1.8  and 
11.5  percent,  respectively)  (McCaughey  1990). 


50- 


?  40- 

a: 

UJ 
(D 

H  30" 
•z. 

UJ 

o 

cc 

lU 
Q. 

lU 
> 


20- 


^  10- 


3 
O 


Burled,  germinating  under  moist  conditions  - 1989 


JUNE 


n- 2.700  seeds 


Buried,  germinating  under  dry  conditions  - 1988 
n-  1,440  seeds 


JULY 


AUGUST    '  SEPTEMBER  '  OCTOBER 


Figure  3 — Cumulative  percent  germination  of  white- 
bark pine  germinants  from  buried  and  protected 
seeds,  sown  in  1987  and  1988  respectively,  germi- 
nating under  dry  (1988)  and  moist  (1989)  conditions. 
Data  collected  from  a  clearcut  at  2,530  m  elevation 
in  southwestern  Montana,  U.S.A. 


182 


Table  1 — Percent  first-year  germination  of  1987  and  1988  sown 

seeds  of  whitebark  pine  as  affected  by  shade  cover,  seed- 
bed condition,  and  sowing  depth.  All  treatments  excluded 
seed  consumers  allowing  for  surface  germination.  Data 
collected  from  a  clearcut  at  2,530  m  elevation  in  south- 
western Montana,  U.S.A. 


First-year  germination 

Factor 

1988 

1989 

Factor 

level 

mean 

mean 

-  Percent  

Shade  cover 

0 

^5.2  (a) 

16.7  (c) 

(percent) 

25 

7.9  (a) 

18.8  (cd) 

50 

6.8  (a) 

20.4  (d) 

Seedbed 

Mineral 

8.2  (a) 

26.0  (c) 

condition 

Litter 

5.1  (b) 

14.5  (d) 

Burned 

15.3  (d) 

Sowing  depth 

Surface 

1.8  (a) 

3.5  (c) 

Buried 

11.5  (b) 

33.7  (d) 

'Similar  and  dissimilar  letters  in  parentheses  within  a  column  for  a  factor 
represent  nonsignificant  and  significant  differences  respectively. 


Delayed  Germination 

Delayed  germinants  are  whitebark  pine  seeds  that  ger- 
minate two  or  more  seasons  after  being  sown  or  cached. 
The  European  stone  pine  (Pinus  cembra)  germinates  in 
the  first,  second,  and  third  year  after  caching  (Krugman 
and  Jenkinson  1974).  Whitebark  pine  seed  germinates 
at  least  3  years  following  sowing  (McCaughey  1990; 
McCaughey,  in  press  b;  McCaughey  and  Schmidt  1990). 
Although  there  were  absolute  value  differences,  the  effects 
of  shade  cover  and  seedbed  condition  became  nonsignifi- 
cant as  germination  occurred  over  a  3-year  period  in  a 
study  of  whitebark  germination  in  western  Montana, 
U.S.A.  (table  2)  (McCaughey,  in  press  b).  Germination 
from  buried  seed  (56  percent)  remained  significantly  higher 
than  for  surface-sown  seed  (7  percent)  3  years  after  sow- 
ing (McCavighey,  in  press  b). 

Dormancy  of  whitebark  pine  seed  is  caused  by  embryo 
underdevelopment,  physiological  embryo  dormancy,  im- 
perviousness  of  the  seed  coat  and  female  gametophyte 
tissue  to  oxygen  and  water  uptake,  and  possibly  by  depo- 
sition of  growth  inhibitors  to  the  embryo  by  female  game- 
tophyte tissue  (Pitel  1981;  Pitel  and  Wang  1980,  1990). 
Seeds  of  whitebark  pine  must  mature  in  a  short  growing 
season  in  high-elevation  forests,  and  if  climatic  conditions 
slow  the  maturation  process  nutcrackers  will  harvest  im- 
mature seed.  Nutcrackers  are  normally  not  concerned 
with  seed  maturity  and  will  also  harvest  seeds  early  in 
the  season  before  they  are  mature.  Embryo  vmderdevel- 
opment  can  be  overcome  by  exposing  imbibed  seeds  to 
20  °C  for  30  to  60  days  (Leadem  1985).  Clipping  the  seed- 
coat  is  a  method  to  overcome  physiological  barriers  to  ger- 
mination (Pitel  1981). 

Under  natural  conditions,  the  greatest  germination 
from  cached  whitebark  pine  seeds  occurs  in  the  second 


year  following  caching  due  to  delayed  dormancy  mecha- 
nisms (McCaughey  1990,  1992;  Tomback  1992).  For 
example,  germination  of  buried  whitebark  pine  seed  in- 
creased from  11  percent  in  the  first  year  following  sowing 
to  45  percent  in  the  second  year.  Germination  declined  to 
11  percent  in  the  third  year  foUovnng  sowing  (McCaughey, 
in  press  b). 

Limiting  Factors 

Germination  rates  for  whitebark  pine  are  highest  from 
mid-June  through  the  end  of  July  (McCaughey  1990). 
Mortality  of  first-year  seedlings  follows  the  same  pattern; 
the  highest  mortality  rates  occvir  when  germination  is 
highest  (fig.  4)  (McCaughey  1990).  Germination  exceeds 
mortality  up  to  the  first  of  August,  resulting  in  an  accu- 
mulation of  surviving  seedlings;  however,  germination 
and  mortality  rates  decrease  after  the  first  of  August. 
Germination  continues  from  early  August  until  the  first 
of  September  with  total  nimabers  of  survivors  remaining 
constant  because  mortality  equals  germination  (McCaughey 
1990).  Late  germinants  have  the  same  probability  of  svir- 
vival  as  early  germinants.  Factors  limiting  seedling  sur- 
vival include: 

Microsite  and  Biotic — Three  major  factors  are: 
(1)  insolation  (heat  scorching  of  seedling  stem  at  ground 
surface),  (2)  drought  (drying  out  of  seedling),  and  (3)  ani- 
mals (burial,  uprooting,  or  nipping),  specifically  pocket 
gophers  (Thomomys  talpoides)  (McCaughey  and  Schmidt 
1990).  Hutchins  and  Lanner  (1982)  observed  a  chipmunk 
uproot  and  consume  a  whitebark  pine  seedling.  Insola- 
tion mortality  of  whitebark  pine  was  higher  on  mineral, 
litter,  and  burned  seedbeds  on  nonshaded  plots  when 
compared  to  25  and  50  percent  shaded  plots  in  a  germina- 
tion test  in  southwestern  Montana,  U.S.A.  (McCaughey 
1990). 

Blister  Rust — The  introduced  disease,  blister  rust 
{Cronartium  ribicola),  poses  the  most  serious  threat  to 


Table  2 — Percent  germination'  of  1 987  sown  whitebark  pine  seed 
(buried  2  to  4  cm)  as  affected  by  shade  cover  and  seed- 
bed condition  for  each  of  the  first  3  years  following  sowing 
and  the  3-year  total.  Data  collected  from  a  clearcut  at 
2,530  m  elevation  in  southwestern  Montana,  U.S.A. 


Emergence 


Factor 


Factor 
level 


First 
year 


Second 
year 


Third 
year 


Three-year 
total 


Shade  cover 
(percent) 

Seedbed 
condition 


0 

25 
50 

Mineral 
Litter 


10 
13 
12 

13 
9 


Percent 

39 
51 
46 

40 
50 


5 
14 
15 

7 
15 


247  (a) 
61  (a) 
60  (a) 

51  (a) 
60  (a) 


'Percent  germination  for  each  year  is  based  on  the  number  of  seeds  that 
had  not  previously  emerged. 

^Similar  and  dissimilar  letters  in  parentheses  for  a  factor  represent  statisti- 
cally nonsignificant  and  significant  differences  respectively. 


183 


10 
8 

6  -I 


1988 


INSOLATION 
DROUGHT 


I 


JUNE 


JULY        AUGUST  SEPTEMBER 


1989 

■1  INSOLATION 

m  DROUGHT 

□  ANIMAL 

JUNE  1 

JULY     1  AUGUST  1  SEPTEMBER 

Figure  4 — Percent  first-year  mortality  of  whitebark 
pine  seedlings  by  causal  agent  over  time  for  1 987 
and  1 988  sown  seeds.  Data  collected  from  a  clear- 
cut  at  2,530  m  elevation  in  southwestern  Montana, 
U.S.A. 


The  Yellowstone  environment  is  poorly  suited  to  blister 
rust  infection  because  it  has  a  cool,  dry  climate  that  is 
only  marginally  suitable  for  blister  rust  teliospore  germi- 
nation (Krebill  1971). 

Mountain  Pine  Beetle — Mountain  pine  beetle  is  the 
most  damaging  insect  in  mature  stands  of  whitebark  pine 
(Arno  and  Hoff  1989;  McCaughey  and  Schmidt  1990). 
Large  numbers  of  whitebark  pine  are  killed  by  moimtain 
pine  beetle  when  epidemics  spread  into  whitebark  pine 
stands  from  the  lower  elevation  lodgepole  pine  zone.  These 
epidemics,  like  blister  rust,  create  conditions  where  suc- 
cession progresses  toward  dominance  by  shade-tolerant 
conifers  such  as  subalpine  fir  and  Engelmann  spruce  (Arno 
1986). 

Fire  Suppression — Whitebark  pine  is  serai  to  subal- 
pine fir  and  other  conifers,  and  periodic  fires  have  helped 
to  perpetuate  the  pine  (Arno  1986).  Past  fire  intervals 
in  whitebark  pine  stands  ranged  from  50  to  300  years  or 
more  depending  on  site  conditions  (Arno  1980).  Fires  oc- 
cur in  whitebark  pine  stands  only  under  severe  burning 
conditions  and  are  tj^ically  non-stand  replacing.  Fires 
sweep  across  large  areas  of  forest  reducing  competition 
from  shade-tolerant  conifers  (Arno  1976).  Nearly  90  years 
of  fire  suppression  have  reduced  the  annual  acreage  of 
whitebark  pine  forests  being  biu'ned,  creating  conditions 
where  shade-tolerant  species  have  taken  over  what  were 
once  serai  whitebark  pine  stands. 

Severe  fire  conditions  are  being  created  by  increased 
mortality  and  the  buildup  of  fuels  in  whitebark  pine  stands 
due  to  mountain  pine  beetle  and  blister  rust  epidemics. 
Stand  replacement  fires  will  be  the  result  of  this  unnatu- 
ral fuel  loading  and  will  eliminate  large  cone-producing 
stands  of  whitebark  pine. 


the  survival  of  whitebark  pine  in  parts  of  its  distribu- 
tional range.  Blister  rust  was  brought  to  western  North 
America  in  1910  on  a  boatload  of  blister  rust-infected 
eastern  white  pine  (Pinus  strobus  L.).  Whitebark  pine  is 
the  most  susceptible  to  blister  rust  infection  of  14  white 
pines  rated  by  Bingham  (1972).  Blister  rust  causes  exten- 
sive damage  and  mortality  in  whitebark  pine  stands  in 
moist  mountain  regions  in  northwestern  Montana,  north- 
ern Idaho,  and  the  Washington  Cascades  in  the  United 
States  (Arno  1986).  Damage  is  light  where  environmental 
conditions  are  somewhat  dry  and  cold,  and  there  is  a  de- 
crease in  low-elevation  sources  of  inoculimi  (Hoff  and 
Hagle  1990). 

Blister  rust-infected  trees  may  live  for  several  years 
before  mortality  occurs;  during  that  time  cone  production 
is  reduced  due  to  loss  of  upper  crowns.  Mortality  from 
rust  favors  succession  toward  dominance  by  whitebark 
pine's  shade-tolerant  associates  thus  reducing  site  poten- 
tial for  whitebark  pine  regeneration  and  survival  (Arno 
1986;  Arno  and  Hoff  1989;  McCaughey  and  Schmidt  1990). 

The  degree  of  blister  rust  infection  on  whitebark  pine 
decreases  southward  for  all  parts  of  its  range  (Hoff  and 
Hagle  1990;  Keane  and  others  1990).  A  major  population 
of  whitebark  pine  exists  in  the  Yellowstone  ecosystem  of 
southwestern  Montana  and  northwestern  Wyoming,  U.S.A. 


VEGETATIVE  REPRODUCTION 

Whitebark  pine  can  reproduce  vegetatively  through  lay- 
ering of  lower  branches  along  the  ground  surface.  In  the 
Mission  Range  of  western  Montana  vegetative  reproduc- 
tion was  observed  from  layering  of  shrublike  (krummholz) 
whitebark  pine  (Arno  1981;  Arno  and  Hoff  1989).  Although 
layering  is  possible,  the  vast  majority  of  reproduction  is 
from  seeds. 

SUMMARY 

Regeneration  processes  are  similar  for  whitebark  pine 
and  the  other  stone  pines  of  the  world.  The  initiation, 
development,  and  maturation  of  cones  and  seeds  for  white- 
bark pine  are  characteristic  of  five-needle  pines  in  general 
with  seasonal  timing  of  these  processes  varying  between 
species.  The  entire  process  from  cone  initiation  to  cone 
and  seed  maturity  takes  about  24  months.  Cone  and  seed 
development  are  most  affected  by  variations  in  climatic 
conditions  during  periods  of  reproductive  bud  formation, 
pollination,  and  growth  and  development.  Several  insects 
and  diseases  reduce  survival  of  cones  and  seeds  of  white- 
bark pine. 

Regeneration  of  whitebark  pine  depends  almost  exclu- 
sively on  the  seed  selection,  dispersal,  and  caching  habits 


184 


of  the  Clark's  nutcracker.  Other  stone  pines  of  the  world 
have  similar  mutualistic  relationships  with  bird  species 
for  seed  dispersal. 

Whitebark  pine  seeds  are  cached  by  Clark's  nutcrackers 
on  a  variety  of  sites  ranging  from  forest  litter  to  cracks 
and  fissures  in  rocks.  The  majority  of  regeneration  occurs 
on  south  or  west  aspects.  Seed  storage  in  south-aspect 
windblown  sites  ensures  that  some  caches  are  snow  free 
in  winter  and  spring  for  nutcracker  retrieval.  Regenera- 
tion densities  are  highest  on  north  aspects  even  though 
fewer  seeds  are  cached  there  because  of  favorable  micro- 
site  conditions  such  as  moistxu'e  and  insolation  protection. 

Germination  of  whitebark  pine  begins  about  mid-June 
and  continues  through  early  September  in  the  first  year 
following  caching.  Delayed  germination  occurs  for  up 
to  3  or  4  years  after  caching.  Total  germination  over  a 
3-year  period  is  highest  on  litter  seedbeds  that  have  50 
percent  shade  cover,  while  germination  is  lower  on  min- 
eral and  nonshaded  seedbeds  (McCaughey,  in  press  b). 
Dormancy  of  whitebark  pine  seed  is  caused  by  embryo 
underdevelopment,  physiological  embryo  dormancy,  and 
imperviousness  of  the  seed  coat  and  female  gametoph3^e 
tissue  to  oxygen  and  water  uptake. 

Factors  limiting  survival  of  whitebark  pine  seedlings 
are  microsite,  biotic,  disease,  insects,  and  fire.  Insolation 
and  drought  are  microsite-associated  factors  that  cause 
seedling  mortality.  Animals  such  as  pocket  gophers  bury, 
uproot,  or  nip  off  young  seedlings,  and  chipmunks  uproot 
and  consimie  whole  seedlings.  Deer  mice  and  southern 
red-backed  voles  and  other  small  mammals  consume  seed 
before  they  germinate.  Blister  rust  causes  mortality  or 
extensive  damage  of  the  cone-bearing  portions  of  the  tree. 
Whitebark  pine  is  killed  by  the  mountain  pine  beetle 
when  epidemics  spread  into  whitebark  pine  stands  from 
the  lower  elevation  lodgepole  pine  zone.  Fire  suppression 
creates  conditions  where  shade-tolerant  species  such  as 
Engelmann  spruce  and  subalpine  fir,  normally  killed  by 
light  imderburns,  take  over  through  succession  what  were 
once  serai  whitebark  pine  stands.  Severe  fire  conditions 
are  being  created  by  the  buildup  of  fuels  in  whitebark 
pine  stands  due  to  increased  mortality. 

We  are  beginning  to  piece  together  the  puzzle  of  the  re- 
generation process  of  whitebark  pine.  Survival  of  wildlife 
species  such  as  the  grizzly  is  threatened  due  to  a  reduc- 
tion or  loss  of  localized  populations  of  whitebark  pine 
caused  by  global  climate  change,  introduced  disease,  in- 
sects, and  succession  fi-om  fire  exclusion.  Future  research 
should  define  microsite  and  biotic  factors  influencing  ger- 
mination and  long-term  survival  of  seedlings.  Informa- 
tion on  regeneration  processes  from  other  stone  pine  spe- 
cies from  around  the  world  is  helping  identify  key  areas 
for  research  on  whitebark  pine.  For  example,  little  infor- 
mation is  available  on  the  elevational  distribution  of 
whitebark  and  which  habitats  are  best  suited  for  regen- 
eration. Cembra  pine  {Pinus  cembra)  grows  in  the  Swiss 
Alps  fi-om  valley  bottoms  to  timberline  when  management 
actions  reduce  intertree  competition.  Whitebark  pine 
management  wiU  improve  from  increased  knowledge  of 
regeneration  processes  of  all  the  stone  pines. 


REFERENCES 

Amo,  S.  F.  1976.  The  historical  role  of  fire  on  the  Bitter- 
root  National  Forest.  Res.  Pap.  INT-187.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Forest  and  Range  Experiment  Station.  30  p. 

Arno,  S.  F.  1980.  Forest  fire  history  in  the  Northern 
Rockies.  Journal  of  Forestry.  78(8):  460-465. 

Arno,  S.  F.  1981.  [Unpublished  observations].  Missoula, 
MT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermoimtain  Fire  Sciences  Laboratory. 

Amo,  S.  F.  1986.  Whitebark  pine  cone  crops  -  a  diminish- 
ing source  of  wildlife  food?  Western  Journal  of  Applied 
Forestry.  1(3):  92-94. 

Arno,  S.  F.;  Hoff,  R.  J.  1989.  Silvics  of  whitebark  pine 
(Pinus  albicaulis).  Gen.  Tech.  Rep.  INT-253.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermoimtain  Research  Station,  lip. 

Bartos,  D.  L.;  Gibson,  K.  E.  1990.  Insects  of  whitebark 
pine  with  emphasis  on  mountain  pine  beetle.  In: 
Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
171-178. 

Bingham,  R.  T.  1972.  Taxonomy,  crossability,  and  relative 
blister  rust  resistance  of  5-needled  white  pines.  In:  Biol- 
ogy of  rust  resistance  in  forest  trees:  proceedings  of  a 
NATO-IUFRO  advance  study  institute;  1969  August 
17-24;  Moscow,  ID.  Misc.  Publ.  1221.  Washington,  DC: 
U.S.  Department  of  Agriculture:  271-280. 

Bock,  W.  J.;  Balda,  R.  P.;  Vander  Wall,  S.  B.  1973.  Mor- 
phology of  the  sublingual  pouch  and  tongue  muscula- 
ture in  Clark's  nutcracker.  Auk.  90:  491-519. 

Craighead,  J.  J.;  Sumner,  J.  S.;  Scaggs,  G.  B.  1982.  A  de- 
finitive system  for  analysis  of  grizzly  bear  habitat  and 
other  wilderness  resources.  Wildlands  Inst.  Monogr.  1. 
Missoula,  MT:  University  of  Montana.  279  p. 

Dewey,  Jerald  E.  1989.  [Personal  communication  to  Dale 
Bartos  of  Intermountain  Research  Station,  Logan,  UT]. 
March  27.  Missoula,  MT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Northern  Region. 

Edwards,  D.  G.  W.  1990.  Cone  prediction,  collection  and 
processing.  In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
78-102. 

Eggers,  D.  E.  1986.  Management  of  whitebark  pine  as  po- 
tential grizzly  bear  habitat.  In:  Contreras,  G.  P.;  Evans, 
K.  E.,  comps.  Proceedings — grizzly  bear  habitat  sympo- 
sium; 1985  April  30-May  2;  Missoula,  MT.  Gen.  Tech. 
Rep.  INT-207.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
170-175. 


185 


Furnier,  G.;  Knowles,  P.;  Clyde,  M.;  Dancik,  B.  1987.  Ef- 
fects of  avian  seed  dispersal  on  the  genetic  structure 
of  whitebark  pine  populations.  Evolution.  41:  607-612. 

Harlow,  W.  M.  1931.  The  identification  of  the  pines  of  the 
United  States,  native  and  introduced,  by  needle  struc- 
ture. Tech.  Publ.  32.  Syracuse,  NY:  New  York  State 
College  of  Forestry.  21  p. 

Hoff,  R.  J.  1992.  Unpublished  data  on  file  at:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station,  Forestry  Sciences  Laboratory,  Moscow, 
ID. 

Hoff,  R.  J.;  Hagle,  S.  1990.  Diseases  of  whitebark  pine 
with  special  emphasis  on  white  pine  blister  rust.  In: 
Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
179-190. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Jacobs,  J.  S.  1989.  Temperature  and  light  effects  on  seed- 
ling performance  of  Pinus  albicaulis.  Bozeman,  MT: 
Montana  State  University.  39  p.  Thesis. 

Jacobs,  J.;  Weaver,  T.  1990a.  Effects  of  temperature  and 
temperature  preconditioning  on  seedling  performance 
of  whitebark  pine.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  white- 
bark pine  ecosystems:  ecology  and  management  of  a 
high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain 
Research  Station:  134-139. 

Jacobs,  J.;  Weaver,  T.  1990b.  Occurrence  of  multiple 
stems  in  whitebark  pine.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-moimtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  156-159. 

Keane,  R.  E.;  Arno,  S.  F.;  Brown,  J.  K.;  Tomback,  D.  F. 
1990.  Simulating  disturbances  and  conifer  succession 
in  whitebark  pine  forests.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  274-288. 

Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  grizzly  and  black 
bears  in  the  Yellowstone  area.  In:  International  Confer- 
ence on  Bear  Research  and  Management.  Madison,  WI: 
International  Bears  Association.  5:  166-173. 

Knight,  R.  R.;  Blanchard,  B.  M.;  Mattson,  D.  J.  1987. 
Yellowstone  grizzly  bear  investigation:  report  of  the 
interagency  study  team,  1987.  Bozeman,  MT:  U.S.  De- 
partment of  the  Interior,  National  Park  Service.  80  p. 


Krebill,  R.  G.  1971.  Conditions  for  teliospore  germination 
in  Yellowstone  National  Park  environment.  Unpub- 
lished report  on  file  at:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Research  Station, 
Ogden,  UT. 

Krugman,  S.  L.;  Jenkinson,  J.  L.  1974.  Pinus  L.  pine. 
In:  Schopmeyer,  C.  S.,  ed.  Seeds  of  woody  plants  in  the 
United  States.  Agric.  Handb.  450.  Washington,  DC: 
U.S.  Department  of  Agriculture,  Forest  Service:  598-638. 

Landis,  T.  D.;  Tinus,  R.  W.;  McDonald,  S.  E.;  Barnett, 
J.  P.  1990.  The  biological  component:  nursery  pests  and 
mycorrhizae.  Vol.  5.  The  container  tree  nursery  manual. 
Agric.  Handb.  674.  Washington,  DC:  U.S.  Department 
of  Agriculture,  Forest  Service.  171  p. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in 
the  ecology  and  evolution  of  limber  and  whitebark 
pines.  In:  Sixth  North  American  forest  biology  work- 
shop proceedings;  1980  August;  Edmonton,  AB. 
Edmonton,  AB:  University  of  Alberta:  14-48. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for 
seed  dispersal  by  Clark's  nutcracker.  Canadian  Journal 
of  Forest  Research.  12(2):  391-402. 

Lanner,  R.  M.;  Vander  Wall,  S.  B.  1980.  Dispersal  of  lim- 
ber pine  seed  by  Clark's  nutcracker.  Journal  of  For- 
estry. 78(10):  637-639. 

Leadem,  C.  L.  1985.  Seed  dormancy  in  three  Pinus  species 
of  the  Inland  Mountain  West.  In:  Schmidt,  Wyman  C, 
comp.  Proceedings — future  forests  of  the  Moimtain 
West:  a  stand  culture  symposium;  1986  September  29- 
October  3;  Missoula,  MT.  Gen.  Tech.  Rep.  INT-243. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station:  117-124. 

Linhart,  Y.;  Tomback,  D.  F.  1985.  Seed  dispersal  by  nut- 
crackers causes  multi-trunk  growth  form  in  pines. 
Oecologia.  67:  107-110. 

McCaughey,  W.  W.  1990.  Biotic  and  microsite  factors 
affecting  Pinus  albicaulis  establishment  and  survival. 
Bozeman,  MT:  Montana  State  University.  78  p. 
Dissertation. 

McCaughey,  W.  W.  1988.  Unpublished  data  on  file  at: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station,  Forestry  Sciences  Labora- 
tory, Bozeman,  MT. 

McCaughey,  W.  W.  1992.  Unpublished  data  on  file  at: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station,  Forestry  Sciences  Labora- 
tory, Bozeman,  MT. 

McCaughey,  W.  W.  [In  press  a].  Seasonal  maturation  of 
whitebark  pine  seed  in  the  Greater  Yellowstone  ecosys- 
tem. In:  Schullery,  Paul;  Despain,  Don,  comps.  Proceed- 
ings of  the  1st  biennial  scientific  conference  on  the 
Greater  Yellowstone  Ecosystem — plants  and  their  envi- 
ronments; 1991  September  16-17;  Mammoth  Hot 
Springs,  Yellowstone  National  Park,  WY.  Tech.  Rep. 
USDI  National  Park  Service. 

McCaughey,  W.  W.  [In  press  b].  Delayed  germination  and 
seedling  emergence  of  Pinus  albicaulis  in  a  high  eleva- 
tion clearcut  in  Montana,  U.S.A.  In:  Edwards,  D.  G., 
comp.  Proceedings — symposium:  seed  dormancy  and 
barriers  to  germination;  1991  April  23-26;  lUFRO 


186 


project  group  P2. 04-00  meeting  on  seed  problems; 
Victoria,  BC.  Victoria,  BC:  Forestry  Canada. 

McCaughey,  W.  W.;  Amo,  S.  F.;  Tomback,  D.  F.  1990. 
Unpublished  personal  observations  from  Henderson 
Mountain  near  Cooke  City,  MT,  U.S.A..  On  file  at:  U.S. 
Department  of  Agriciilture,  Forest  Service,  Intermoun- 
tain  Research  Station,  Forestry  Sciences  Laboratory, 
Bozeman,  MT. 

McCaughey,  W.  W.;  Schmidt,  W.  C.  1990.  Autecology  of 
whitebark  pine.  In:  Schmidt,  Wjonan  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermoimtain  Research 
Station:  85-96. 

Mirov,  N.  T.  1946.  Viability  of  pine  seed  after  prolonged 
cold  storage.  Journal  of  Forestry.  44(3):  193-195. 

Mirov,  N.  T.  1967.  The  genus  Pinus.  New  York:  The  Ronald 
Press.  602  p. 

Morgan,  Penny;  Bunting,  Stephen  C.  1990.  Fire  effects 
in  whitebark  pine  forests.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposiimi 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  166-170. 

Pfister,  R.  D.;  Kovalchik,  B.  L.;  Arno,  S.  F.;  Presby,  R.  C. 
1977.  Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  174  p. 

Pitel,  J.  A.  1981.  Improved  germination  of  whitebark 
pine  {Pinus  alhicaulis  Engelm.)  seeds.  Chalk  River,  ON: 
Canadian  Forestry  Service,  Petawawa  National  For- 
estry Institute;  unpublished  report.  5  p. 

Pitel,  J.  A.;  Wang,  B.  S.  P.  1980.  A  prehminary  study  of 
dormancy  in  Pinus  albicaulis  seeds.  Bi-Monthly  Re- 
search Notes.  36(1):  4-5. 

Pitel,  J.  A.;  Wang,  B.  S.  P.  1990.  Physical  and  chemical 
treatments  to  improve  germination  of  whitebark  pine 
seeds.  In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-movmtain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agriciil- 
ture,  Forest  Service,  Intermoimtain  Research  Station: 
130-133. 

Reinhart,  D.  P.  1990.  [Personal  commmiication].  August. 
Bozeman,  MT:  U.S.  Department  of  the  Interior; 
Montana  State  University,  Interagency  Grizzly  Bear 
Project. 

Reinhart,  D.  P.;  Mattson,  D.  J.  1990.  Red  sqiiirrels  in  the 
whitebark  zone.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
moxmtain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermoimtain  Research 
Station:  256-263. 


Schmidt,  W.  C;  Lotan,  J.  E.  1980.  Phenology  of  common 
forest  flora  of  the  Northern  Rockies— 1928  to  1937.  Res. 
Pap.  INT-259.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Forest  and  Range 
Experiment  Station.  20  p. 

Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  1990. 
Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station. 
386  p. 

Schubert,  G.  H.  1954.  Viability  of  various  coniferous  seeds 
after  cold  storage.  Journal  of  Forestry.  52:  446-447. 

Smith,  C.  C.  1968.  The  adaptive  nature  of  social  organiza- 
tion in  the  genus  of  tree  squirrels  Tamiasciurus .  Eco- 
logical Monographs.  38:  31-63. 

Steele,  R.;  Cooper,  S.  V.;  Ondov,  D.  M.;  Roberts,  D.  W.; 
Pfister,  R.  D.  1983.  Forest  habitat  types  of  eastern 
Idaho-western  Wyoming.  Gen.  Tech.  Rep.  INT-144. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Forest  and  Range  Experiment 
Station.  122  p. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. Living  Bird.  16(1977):  123-160. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51(2):  451-467. 

Tomback,  D.  F.  1983.  Nutcrackers  and  pines:  coevolution 
or  coadaptation?  In:  Nitecki,  M.  H.,  ed.  Coevolution. 
Chicago  and  London:  University  of  Chicago  Press: 
179-223. 

Tomback,  D.  F.  1986.  Post-fire  regeneration  of  krummholz 
whitebark  pine:  a  consequence  of  nutcracker  seed  cach- 
ing. Madrono.  33(2):  100-110. 

Tomback,  D.  F.  1992.  Unpublished  data  on  file  at:  Depart- 
ment of  Biology,  University  of  Colorado  at  Denver, 
Denver,  CO. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Co- 
evolution  of  whitebark  pine  and  nutcrackers:  implica- 
tions for  forest  regeneration.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposiima 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  118-129. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of 
the  Clark's  nutcracker  and  the  pinon  pine  for  efficient 
seed  harvest  and  dispersal.  Ecological  Monographs. 
47:  89-111. 

Weaver,  T.;  Jacobs,  J.  1990.  Occurrence  of  multiple  stems 
in  whitebark  pine.  In:  Schmidt,  Wjonan  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  white- 
bark pine  ecosystems:  ecology  and  management  of  a 
high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station:  156-159. 


187 


JAPANESE  STONE  PINE  CONE 
PRODUCTION  ESTIMATED  FROM  CONE 
SCARS,  MOUNT  KISOKOMAGATAKE, 
CENTRAL  JAPANESE  ALPS 

Ikuko  Nakashinden 


Abstract — The  past  cone  production  of  the  Japanese  stone  pine 
(Pinus  pumila)  on  Mount  Kisokomagatake  was  reconstructed 
from  cone  scars.  The  main  cone  production  was  by  larger  indi- 
viduals in  the  lower  part  of  the  scrub  zone;  it  was  greater  than 
that  in  the  upper  part  where  growth  and  crop  are  controlled 
mainly  by  strong  winter  winds  and  snow. 


The  Japanese  stone  pine  {Pinus  pumila  Regal)  is  a 
genetic  dwarf  pine  distributed  over  the  eastern  part  of 
Siberia  and  China,  the  northern  part  of  Korea,  and  Japan. 
In  Japan,  most  of  its  habitats  are  located  on  high  moun- 
tains. The  climax  scrub  zone  is  formed  above  the  timber- 
line  of  subalpine  coniferous  forests.  At  the  southern  mar- 
gin of  its  range  (Chubu  District)  the  scrub  zone  occupies 
the  altitudinal  zone  from  1,961  to  3,192  m  (Yanagimachi 
and  Ohmori  1991). 

The  Japanese  stone  pine  propagates  in  two  different 
ways:  one  by  the  expansion  of  shoots  and  the  other  by 
seeds.  The  latter  is  the  most  important  for  spatial  ex- 
pansion (Okitsu  1990).  Seeds  are  dispersed  by  seed- 
storing  corvids  {Nucifraga  caryocatactes)  (Hayashida 
1989;  Tomback  1990).  Studies  concerning  pine  cone 
production  have  begun  recently.  Okitsu  and  Mizoguchi 
(1990,  1991)  evaluated  pine  cone  productivity  based  on 
the  individual  growth  sizes  and  scrub  sizes.  In  addition, 
studies  of  the  relationship  between  environment  of 
the  habitats  and  cone  production  have  commenced 
(Nakashinden  1990).  These  studies  are  limited  to  1  year's 
cone  production  data.  It  is  notable  that  past  cone  produc- 
tion can  be  estimated  from  cone  scars.  This  was  done  on 
whitebark  pine  {Pinus  albicaulis)  in  the  United  States 
(Morgan  and  Bimting  1992;  Weaver  and  Forcella  1986). 
Saito  and  others  (1989)  reported  that  cone  scars  are  also 
recognizable  in  the  Japanese  stone  pine  in  Hokkaido,  but 
little  information  about  it  was  given  in  their  report. 

This  author  developed  the  method  to  reveal  the  yearly 
change  in  cone  production  for  the  Japanese  stone  pine  and 
succeeded  in  reconstructing  pine  cone  production  over  the 
past  15  years  (Nakashinden  1991).  Using  this  cone  scar 
method,  chronological  pine  cone  production  of  the  Japa- 
nese stone  pine  is  described  and  the  relationship  to  the 
environment  of  its  habitats  is  examined. 


STUDY  AREA  AND  METHODS 

The  study  area  is  located  on  Moimt  Kisokomagatake, 
whose  simimit  is  2,956  m,  in  the  Central  Japanese  Alps, 
in  Chubu  District  (fig.  1).  This  location  is  one  of  the 
southernmost  parts  of  the  Japeinese  stone  pine 
distribution. 

The  habitats  examined  are  located  on  the  mountain 
slopes  between  2,600  and  2,950  m  elevation,  including 
the  summits  of  Mount  Kisokomagatake  and  Mount 
Chausudake  (2,652  m).  In  this  altitudinal  zone,  the 
Japanese  stone  pine  is  dominant,  forming  its  own  vege- 
tation zone  above  the  subalpine  forest  of  Abies  veitchii, 
Tsuga  diversifolia,  Betula  ermanii,  and  Alnus 
maximowiczii. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Ikuko  Nakashinden  is  Graduate  Student,  Department  of  Geography, 
University  of  Tokyo,  Bunkyo-ku,  Tokyo  113,  Japan. 


Figure  1— Location  of  Mount  Kisokomagatake. 


188 


Pine  Cone  Scars 

The  positions  of  pine  cones  of  the  Japanese  stone  pine 
are  restricted  to  the  first  nodes  of  the  annual  shoots. 
They  matiire  in  the  second  year  after  fruiting.  In 
September,  nutcrackers  harvest  cones  of  the  Japanese 
stone  pine  as  a  food  source.  When  the  pine  cone  is  picked, 
a  scar  of  the  cone  remains  on  the  stem  (Nakashinden 
1991;  Saito  and  others  1989).  Each  scar  represents  one 
cone.  From  these  scars,  the  number  of  pine  cones  can  be 
calciilated.  The  production  can  be  determined  from  the 
location  of  the  nodes  on  the  stems.  This  cone  scar  method 
was  used  on  whitebark  pine  in  the  United  States,  and  the 
scars  could  be  measured  for  6  to  8  years  (Weaver  and 
Forcella  1986),  or  6  to  12  years  (Morgan  and  Bunting 
1992).  For  the  Japanese  stone  pine,  the  scars  could  be 
counted  for  15  to  20  years  (Nakashinden  1991). 

Using  this  method,  32  plots  (1  m^)  at  different  altitudes 
from  2,600  m  near  the  forest  hne  to  2,950  m  near  the 
summit  were  investigated.  Ten  stems  in  each  plot  were 
selected  for  counting  pine  cone  scars  and  for  the  measure- 
ment of  individual  stem  growth  for  the  15  years  preceding 
1990. 

Distribution,  Geomorphology,  and 
Snow 

Using  color  aerial  photographs  and  field  observation, 
the  Japanese  stone  pine  scrubs  in  the  study  area  were 
delineated  and  mapped.  Some  topographical  maps  were 
made  during  the  field  survey  in  order  to  examine  the  mi- 
croscale  landform  beneath  the  scrubs.  The  snow  melting 
status  and  snow  melting  period  near  the  simimit  of 
Moxmt  Kisokomagatake  was  observed  and  recorded  on 
photographs  in  each  spring  season  from  1989  to  1991. 


250  - 

200  - 


0) 

o 

0 


150  - 


100  - 


50- 


r— T 

76  77  78  79  80  81  82 

Year 


II   I  '  I   I  I  I 

83  84  85  86  87  88  89  90 


Figure  2 — Annual  cone  crops  of  Japanese  stone 
pine  (1976-90).  Cone  crops  were  reconstructed 
from  cone  scars  on  the  stems.  Total  number  from 
32  plots  is  shown  for  each  year. 


V) 

o 
o 
E 

C 

o 
O 


4.0 


3.0 


2.0 


1.0 


1983 


J  L 


J  L 


76 


78 


80 


82  84 

Year 


86  88 


90 


Figure  3 — Mean  annual  stem  elongation  of 
Japanese  stone  pine  (1976-90).  Ten  stems 
each  in  32  plots  were  measured. 


CONE  PRODUCTION  OVER  15  YEARS 

The  past  15  years  of  pine  cone  production  was  recon- 
structed on  the  study  area  (fig.  2).  Mast  years  were 
recognized  in  1981,  1984,  1988,  and  1989,  showing  a 
3-  or  4-year  periodicity  of  cone  production.  The  lowest 
production  during  the  past  15  years  occurred  in  1982  and 
1983.  In  1983,  the  annual  shoot  length  was  also  the  low- 
est during  the  15-year  period  (fig.  3).  The  year  1983  is 
judged  to  be  one  of  the  worst  years  for  both  cone  crop  and 
growth. 

On  the  individual  plots,  both  mast  and  fail  years  oc- 
curred regardless  of  their  environmental  differences. 
For  the  mast  years,  68  percent  of  the  32  plots  showed 
high  production  in  1981,  and  in  1984  it  was  65  percent. 
For  the  fail  years,  59  percent  of  the  plots  failed  in  cone 
production  in  1982,  and  in  1983  it  was  63  percent.  These 
occurrences  were  dependent  on  the  climate  and  not  on 
the  growth  forms  or  the  location  of  scrubs. 

DISTRffiUTION  PATTERNS  BY 
ALTITUDE 

On  Mount  Kisokomagatake,  the  Japanese  stone  pine 
scrubs  dominate  and  form  a  vertical  zone  above  the  subal- 
pine  forest  at  an  elevation  between  2,600  and  2,950  m. 
However,  the  distribution  patterns  were  characterized  by 
the  altitude.  On  the  upper  part  from  2,850  up  to  2,950  m, 
near  the  summit  of  Moimt  Kisokomagatake,  the  distribu- 
tion pattern  of  the  scrubs  shows  a  stripe  pattern  (fig.  4). 

The  scrubs  are  fragmented  by  huge  granite  rocks,  tors, 
stone  bank  terraces,  trails,  and  alpine  plant  communities 
such  as  Arctous  alpinus  vai.Japonicus,  Diapensia  lap- 
ponia  var.  obovata,  Oxytropis  japonica  Maxim,  Leonto- 
podium  shinanense  Kitam,  Gentiana  algida  Pall.,  and 
Loiseleuria  procumhens  Desv.  The  lower  part  between 
2,850  and  2,600  m,  near  the  boimdary  to  the  subalpine 
forest  shows  an  extensively  spreading  scrub,  covering  the 
slopes  entirely  from  the  ridges  down  to  the  forest  limit 
(fig.  5). 


189 


Figure  4 — Distribution  pattern  of  Japanese  stone 
pine  scrubs  in  tine  upper  part  of  the  Mount 
Kisokomagatake  study  area.  Distribution  is  frag- 
mented by  huge  rocks,  tors,  stone  bank  terraces, 
trails,  and  alpine  plant  communities. 


GROWTH  FORM  AND  CONE 
PRODUCTION  RELATED  TO 
ALTITUDE 

Growth  form  and  pine  cone  production  are  ditTerent  be- 
tween the  upper  and  the  lower  part  of  the  Japanese  stone 
pine  scrub  zone  (table  1).  The  mean  scrub  height,  stem 
length,  and  diameter  at  the  stem  base  in  the  lower  part 
are  almost  twice  as  large  as  those  in  the  upper  part.  The 
mean  pine  cone  production  per  stem  is  1.5  times  more 
than  in  the  upper  part.  Although  the  niunber  of  stems 
per  square  meter  in  the  upper  part  was  twice  that  of  the 
lower  part,  the  Japanese  stone  pine  scrubs  in  the  lower 
part  were  larger  in  size  and  produced  more  cones  than  the 
scrubs  in  the  upper  part. 


DISCUSSION  AND  CONCLUSIONS 

On  Moimt  Kisokomagatake,  the  mean  summer  tem- 
perature (July  and  August)  was  10  to  12  °C  at  2,600  m 
(Koizumi  1974)  and  10  °C  (Nakashinden  1990,  unpub- 
lished) at  2,850  m.  Concerning  the  lower  parts  of  the 
Japanese  stone  pine  scrub  zone,  the  thermal  conditions 
indicate  that  the  Japanese  stone  pine  scrub  zone  is  in- 
cluded in  the  thermal  subalpine  zone  (Yanagimachi  and 
Ohomori  1991).  There,  the  Japanese  stone  pine  scrubs 
are  considered  to  grow  and  produce  cones  at  the  same 
rates  as  in  the  subalpine  zone. 

The  scrubs  in  the  upper  part  are  explained  by  other  en- 
vironmental factors  unfavorable  for  growth  and  cone  pro- 
duction. In  the  Japanese  stone  pine  scrub  zone,  the  scrub 
height  is  related  to  winter  wind  strength  and  snow  depth 
of  their  habitats  (Okitsu  1987).  In  high-moimtain  areas, 
the  Japanese  stone  pine  needs  to  be  protected  by  snow- 
pack  to  evade  physical  damages  or  dehydration  caused  by 
strong  winds.  Short  scrub  height  indicates  strong  winds 
and  shallow  snow  cover  at  the  habitats  during  the  winter 
season,  while  tall  scrub  height  indicates  moderate  winds 
and  heavy  snowpack. 

On  Movmt  Kisokomagatake,  the  depth  of  snow  varies 
from  10  cm  to  3  m  according  to  landforms  and  wind  veloc- 
ity (Koiziuni  1974).  Especially  around  the  simimit,  most 
of  the  snow  is  blown  away  from  the  windward  slope  by 
strong  westerly  winds  (fig.  6).  Snow  accumulates  on  the 
leeward  side  of  the  ridges  or  rocks.  On  the  windward 
slope,  characterized  by  shallow  snow  cover,  the  Japanese 
stone  pine  grow  close  to  the  ground,  thus  protecting  them- 
selves from  strong  winds  (fig.  7).  Therefore,  vertical 
growth  of  the  Japanese  stone  pine  scrubs  is  disturbed  by 
winter  winds. 

Variation  in  snow  depth  caused  differences  in  the 
length  of  the  snow  melting  period.  According  to  3  years' 
observation  by  this  author,  snow  on  the  Japanese  stone 
pine  scrubs  located  on  windward  slopes  started  to  melt 


Figure  5 — Distribution  pattern  of  Japanese 
stone  pine  scrubs  in  the  lower  part  of  the 
Mount  Kisokomagatake  study  area.  They 
spread  extensively  and  cover  the  slopes  en- 
tirely from  the  ridges  down  to  the  forest  line. 


Table  1 — The  growth  and  pine  cone  production  in  each  part  of 
the  distribution  zone  viewed  from  different  altitude  of 
quadrats.  The  boundary  of  the  upper  and  the  lower  part 
is  2,850  m 


Cone  production  and  growth 

Upper  part 

Lower  part 

Cone  production  (pieces) 

4.1 

6.4 

Mean  scrub  height  (cm) 

36.2 

89.2 

Mean  stem  length  (cm) 

92.0 

180.0 

Mean  diameter  at  stem  base  (cm) 

2.2 

4.2 

Mean  annual  stem  elongation  (cm) 

2.4 

3.0 

Number  of  stems/m^ 

38.5 

20.0 

'The  values  indicate  the  means  of  20  plots  for  the  upper  part  and  12  plots 
for  the  lower  part. 


190 


Figure  6 — Mount  Kisokomagatake  summit  in 
early  spring  (1 991 ).  Snow  depth  varies  accord- 
ing to  wind  velocities  and  topography.  The 
western-most  part  (left)  where  snow  cover  was 
thin  due  to  strong  westerly  winds  shows  earliest 
snow  melt  in  spring. 


at  the  end  of  April.  On  the  leev^^ard  slopes,  the  scrubs 
became  completely  free  of  snow  by  the  middle  of  June. 
On  the  slopes  where  snow  remains  imtil  the  end  of  July, 
the  Japanese  stone  pine  is  replaced  by  alpine  snowbed 
vegetation.  On  the  other  hand,  shallow  snow  cover  causes 
early  snow  melt  and  solifluction  is  very  common.  Solifluc- 
tion  forms  stone  banked  terraces  and  cuts  off  the  roots  of 
young  plants.  For  this  reason,  the  Japanese  stone  pine 
scrubs  in  the  upper  part  are  not  able  to  cover  all  the 
slopes  and  thus  show  a  fragmented  distribution  pattern. 

These  severe  environmental  conditions  are  not  favor- 
able for  growth  or  cone  production  of  the  Japanese  stone 
pine.  The  main  cone  production  is  performed  by  large  and 


Figure  7 — Creeping  Japanese  stone  pine  in  the 
upper  part  of  the  study  area.  The  scrubs'  height  is 
10-15  cm.  The  stem  has  branches  flagged  toward 
the  leeward  side.  The  plot  shows  one  of  the  low- 
est cone  productions  (only  one  cone  for  1 0  stems) 
during  these  1 5  years. 


tall  scrubs  at  the  lower  elevations  where  winter  winds  are 
moderate  and  snow  is  evenly  distributed. 

ACKNOWLEDGMENTS 

I  thank  Professor  Dr.  F.-K.  Holtmeier,  Dr.  Wyman  C. 
Schmidt,  and  Dr.  W.  Schonenberger  for  their  invitation 
to  attend  the  workshop.  I  thank  Dr.  Professor  Hiroo 
Ohmori  and  Dr.  Takeei  Koizumi  for  their  valuable  advice 
in  carrying  out  this  study.  My  thanks  also  go  to  Dr.  Tad 
Weaver  and  Dr.  Penelope  Morgan  for  providing  me  with 
information  about  cone  scar  methods  on  whitebark  pine. 
Dr.  Mitsuhiro  Hayashida  offered  me  valuable  information 
in  preparation  for  the  workshop.  I  am  grateful  to  all  my 
colleagues  in  the  Geography  Department  of  Tokyo 
Gakugei  University,  who  enthusiastically  helped  me  in 
my  field  research  on  Mount  Kisokomagatake.  I  thank 
Mr.  Wataru,  Mrs.  Joyce  Mukai,  and  Mr.  Garrett  Vance 
who  checked  my  manuscript.  And  I  thank  Dr.  Miki 
Kobayashi  who  always  encourages  me  to  keep  studying 
Pinus  pumila. 

REFERENCES 

Hayashida,  M.  1989.  Seed  dispersal  and  regeneration  pat- 
terns of  Pinus  parviflora  var.  pentaphylla  on  Mt.  Apoi 
in  Hokkaido.  Research  Bulletins  of  the  College  Experi- 
mental Forests,  Faculty  of  Agriculture,  Hokkaido  Uni- 
versity. 46-1:  177-190. 

Kajimoto,  T.  1989.  Above-groimd  biomass  and  litterfall  of 
Pinus  pumila  scrubs  growing  on  the  Kiso  Mountain 
Range  in  central  Japan.  Ecological  Research.  4:  55-69. 

Koiziuni,  T.  1974.  Landschaftsokologische  Unter- 
suchungen  in  der  Alpine  Stufe  des  Kisokomagatake  in 
der  Japanischen  Zentralalpen  mit  Besonderer  Beriick- 
sichtigung  der  Vegeta  Strukturbodens.  Japanese  Jour- 
nal of  Ecology.  24(2):  78-91. 

Mizoguchi,  T.;  Okitsu,  S.  1987.  Cone  production  of  Pinus 
pumila  Regal  in  high  mountains  in  central  Japan  with 
special  attention  to  basal  diameter  and  annual  elonga- 
tion on  it.  In:  Transactions  of  the  98th  annual  meeting 
of  the  Japanese  Forestry  Society;  1987  April;  K5aishu: 
363-364. 

Morgan,  P.;  Bunting,  S.  C.  1992.  Technical  note:  cone 
crops  of  whitebark  pine.  Western  Journal  of  Applied 
Forestry.  7(3):  71-73. 

Nakashinden,  I.  1990.  Pine  cone  production  of  Pinus 
pumila  and  their  environment  in  Mt.  Kisokomagatake, 
Nagano  prefecture.  In:  Transactions  of  the  38th  autum- 
nal meeting  of  the  Association  of  Japanese  Geogra- 
phers; 1990  September;  Niigata:  104-105. 

Nakashinden,  I.  1991.  Pine  cone  production  of  Pinus 
pumila  reconstructed  from  pine  cone  scars.  In:  Transac- 
tions of  the  39th  spring  meeting  of  the  Association  of 
Japanese  Geographers;  1991  April;  Tokyo:  28-29. 

Okitsu,  S.  1984b.  Comparative  studies  on  the  Japanese 
alpine  zone,  with  special  reference  to  the  ecology  of 
Pinus  pumila  thickets.  Geographical  Review  of  Japan. 
57(11):  170-172. 

Okitsu,  S.  1987.  Pinus  pumila  zone.  In:  Ito,  K.,  comp.  Veg- 
etation of  Hokkaido.  Hokkaido,  Japan:  Hokkaido  Univer- 
sity Press:  129-167. 


191 


Okitsu,  S.;  Ito,  K.  1983.  Dynamic  ecology  of  the  Pinus 
pumila  community  of  Mts.  Taisetsu,  Hokkaido,  Japan. 
The  Environmental  Science,  Hokkaido  University.  6(1): 
151-184. 

Okitsu,  S.;  Mizoguchi,  T.  1986.  Cone  production  of  Pinus 
pumila  community  on  Mt.  Senjo,  Akaishi  Range,  cen- 
tral Japan.  In:  Transactions  of  the  97th  annual  meeting 
of  the  Japanese  Forestry  Society;  1986  April; 
Utsunomiya:  343-344. 

Okitsu,  S.;  Mizoguchi  T.  1990.  Relation  between  cone 
production  and  stem  diameter  and  stem  elongation 
of  Pinus  pumila  of  Japanese  high  mountains.  Japanese 
Journal  of  Ecology.  40(2):  49-55. 

Okitsu,  S.;  Mizoguchi,  T.  1991.  Cone  production  of  Pinus 
pumila  scrubs.  Japanese  Journal  of  Ecology.  41(2): 
101-107. 

Saito,  S.;  Kawabe,  M.;  Nakaoka,  T.  1989.  On  the  forest 
vegetation  of  Mt.  Upepesanke,  Tokachi,  Hokkaido.  (1): 
On  a  thicket  of  Pinus  pumila  at  the  ridge  of  1610  m. 


Bulletin  of  the  Higashi  Taisetsu  Museum  of  Natural 
History.  11:  23-34. 
Saito,  S.;  Kawabe,  M.  1990.  On  the  forest  vegetation  of 
Mt.  Higashi-Nupukaushinupuri,  Tokachi,  Hokkaido. 
(2):  On  two  thickets  of  Pinus  pumila.  Bulletin  of  the 
Higashi  Taisetsu  Museimi  of  Natural  History.  12: 
17-29. 

Tomback,  D.;  Linhart,  Y.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Weaver,  T.;  Forcella,  F.  1986.  Cone  production  in 
Pinus  albicaulis  forests.  In:  Shearer,  R.  C,  comp. 
Proceedings — conifer  tree  seed  in  the  Inland  Mountain 
West;  1985  August  5-6;  Missoula,  MT.  Gen.  Tech.  Rep. 
INT-203.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Research  Station:  68-76. 

Yanagimachi,  O.;  Ohmori,  H.  1991.  Ecological  status  of 
Pinus  pumila  scrub  and  the  lower  boimdary  of  the 
Japanese  alpine  zone.  Arctic  and  Alpine  Research. 
23(4):  242-435. 


192 


EFFECTS  OF  SEED  DISPERSAL  BY 
CLARK'S  NUTCRACKER  ON  EARLY 
POSTFIRE  REGENERATION  OF 
WHITEBARK  PINE 

Diana  F.  Tomback 


Abstract — Clark's  nutcrackers  {Nucifraga  columbiana)  bury 
whitebark  pine  (Pinus  alhicaulis)  seeds  in  recently  burned  ter- 
rain. Studies  of  early  postfire  whitebark  pine  regeneration  in  the 
Northern  Rocky  Mountains  indicate  that  seed  dispersal  by  nut- 
crackers (1)  may  be  more  effective  than  seed  dispersal  by  wind  in 
restocking  large  bums,  particularly  when  the  seed  source  is  unfa- 
vorably positioned  with  respect  to  prevailing  winds,  and  (2)  pro- 
duces a  negative  exponential  distribution  when  whitebark  pine 
regeneration  density  is  plotted  vs.  distance  from  the  seed  source. 
Whitebark  pine  seeds  can  remain  dormant  for  one  or  more  years 
after  dispersal,  which  results  in  continuity  of  regeneration  over 
time. 


1986),  although  updrafts  and  storms  may  disperse  some 
seeds  over  longer  distances.  Often,  the  regeneration  from 
nutcracker  caches  is  in  the  form  of  whitebark  pine  seed- 
Ung  clusters  (for  example  see  Tomback  1982,  1986;  Tomback 
and  others  1990, 1993),  which  may  result  in  mature  multi- 
genet  tree  clusters  (see  Tomback  and  Schuster,  these  pro- 
ceedings, and  references  therein).  In  this  paper  I  summa- 
rize the  results  of  recent  studies  that  compare  patterns  of 
early  postfire  regeneration  of  the  bird-dispersed  white- 
bark pine  and  its  wind-dispersed  forest  associates. 

REGENERATION  IN  WESTERN 
MONTANA 

Studies  of  two  large,  severe  subalpine  bums  in  western 
Montana  (Bitterroot  National  Forest,  Ravalli  County) 
illustrate  how  seed  dispersal  by  nutcrackers  affects  pat- 
terns of  early  postfire  regeneration  of  whitebark  pine  and 
associated  conifers.  These  two  study  areas  were  selected 
for  similarities  in  topography,  severity  of  bum,  forest 
types,  and  time  elapsed  since  fire.  In  this  region  of  west- 
em  Montana,  whitebark  pine  is  abundant  between  about 
2,290  and  2,620  m  elevation  (Amo  and  Hoff  1989;  Pfister 
and  others  1977),  and  the  prevailing  winds  in  late  sum- 
mer and  fall  are  from  the  west  and  southwest  (Finklin 
1983).  The  prefire  forest  in  both  study  areas  was  a  serai 
whitebark  pine  community.  Reported  results  fi-om  these 
studies  are  summarized  here  from  Tomback  and  others 
(1990,  1993). 

The  Sleeping  Child  Bum  resulted  from  a  wildfire  in 
1961  that  bumed  about  11,350  ha  of  forest  on  the  west 
slope  of  the  Sapphire  Range.  The  nearest  whitebark  pine 
seed  sovu*ce  is  mature  forest  above  2,250  m  on  the  eastern 
edge  of  the  bum.  Subalpine  fir,  Engelmann  spmce  (Picea 
engelmannii),  and  lodgepole  pine  (Pinus  contorta)  are  also 
present  in  this  upper  seed  source  as  well  as  around  the 
lower  perimeters  of  the  bum.  Because  the  prevailing 
winds  are  fi*om  the  west,  vdnd-dispersed  seeds  are  irregu- 
larly dispersed  into  the  eastem  portion  of  the  bum  fi'om 
the  upper  seed  source. 

In  1987,  we  established  a  3.6-km-long  transect  along 
an  eastwest  trending  ridge,  running  from  the  seed  soiirce 
into  the  northem  center  of  the  burn  (fig.  2  in  Tomback 
and  others  1990).  Elevation  along  the  ridge  decreases 
from  about  2,480  to  2,150  m  with  increasing  distance 
from  the  eastern  edge  of  the  bum.  Every  150  m  along  the 
transect  we  established  parallel  quadrats,  each  50  m  long, 
on  the  ridgetop,  north,  and  south  aspects  of  the  ridge  for 


Fire  is  an  ecologically  important,  recurring  event  in 
most  western  North  American  montane  forests,  including 
those  regions  where  whitebark  pine  {Pinus  albicaulis)  is 
found.  Fire-prone  areas  are  those  with  fuel  accumulation, 
dry,  windy  weather,  and  periods  of  frequent  lightning 
storms.  In  the  Northem  Rocky  Mountains,  fire  burns 
whitebark  pine  communities  every  50  to  300  years  (Amo 
1980;  Amo  and  Petersen  1983;  Romme  1980).  Stand- 
replacing  fires  result  in  the  renewal  of  successional  white- 
bark pine  communities,  and  underbums  kill  the  less  fire- 
resistant  subalpine  fir  {Abies  lasiocarpa).  In  the  absence 
of  fire  in  these  commimities,  whitebark  pine  is  often  re- 
placed by  more  shade-tolerant  conifers,  particularly  sub- 
alpine fir  (Amo  1986;  Amo  and  Hoff  1989). 

Studies  of  early  postfire  regeneration,  defined  here  as 
the  first  3  decades  after  fire,  provide  information  on  the 
dynamics  of  seed  dispersal  and  seedling  establishment  for 
whitebark  pine  and  associated  conifers.  Clark's  nutcrack- 
ers {Nucifraga  columbiana)  are  the  principal  dispersal 
agents  for  whitebark  pine  seeds,  in  contrast  to  other  coni- 
fers whose  seeds  are  wind-dispersed.  Nutcrackers  bury 
caches  of  one  to  15  or  more  ripe  whitebark  pine  seeds  at 
distances  from  parent  trees  of  a  few  meters  to  22  km 
(Hutchins  and  Lanner  1982;  Tomback  1978, 1982;  Vander 
Wall  and  Balda  1977;  see  review  in  Tomback  and  Linhart 
1990).  In  contrast,  most  vraid-dispersed  seeds  land  only 
within  about  120  m  of  parent  trees  (McCaughey  and  others 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Diana  F.  Tomback  is  Associate  Professor  of  Ecology,  Department  of 
Biology  and  Center  for  Environmental  Sciences,  University  of  Colorado 
at  Denver,  Denver,  CO  80217-3364. 


193 


a  total  of  63  quadrats.  Quadrat  widths  varied  in  preset 
increments  from  1.25  to  15  m,  with  respect  to  local  densi- 
ties of  whitebark  pine.  For  each  quadrat  we  recorded  all 
conifer  regeneration. 

The  Saddle  Moimtain  Burn  resulted  from  a  wildfire  in 
1960  that  spread  over  1,240  ha  of  forest  in  the  Bitterroot 
Mountains  west  of  Lost  Trail  Pass.  Running  northeast 
to  southwest  in  the  longest  dimension,  the  burned  area 
ranges  in  elevation  from  about  1,950  to  2,475  m.  On  the 
southwest  edge  of  the  burn  above  2,275  m,  the  whitebark 
pine  seed  source  is  positioned  so  that  the  burned  area  is 
leeward  with  respect  to  prevailing  winds.  Subalpine  fir, 
Engelmann  spruce,  and  lodgepole  pine  also  occur  in  this 
upper  seed  source  and,  with  Douglas-fir  {Pseudotsuga 
menziesii),  around  the  lower  perimeters  of  the  burn. 

In  1988  we  established  two  parallel  3.2-km-long  tran- 
sects beginning  at  the  seed  source  and  heading  northeast 
through  the  southern  portion  of  the  burn  (fig.  4  in  Tomback 
and  others  1990).  We  set  up  41  quadrats  and  recorded  all 
regeneration  according  to  the  general  methodology  out- 
lined earlier  for  the  Sleeping  Child  Burn  (for  details  of 
methodology,  see  Tomback  and  others  1990,  1993). 

Mean  whitebark  pine  regeneration  density  sampled  per 
quadrat  in  the  Sleeping  Child  Burn  was  much  greater 
than  that  of  its  shade-tolerant  competitors,  subalpine  fir 


Table  1 — Mean  density  and  standard  deviation  (S.D.)  per  quadrat 
for  conifer^  regeneration  26  years  and  28  years  after  fire 
in  tlie  Sleeping  Child  and  Saddle  Mountain  Burns, 
respectively.  Based  on  Tomback  and  others  (1 990) 

Regeneration  sites  per 
ABLA      PIEN       PICO  PSME 


PIAL 


Sleeping  Child 

X  0.070  0.008  0.004  0.039 

S.D.  .104  .014  .009  .066 

Saddle  Mountain 

X  .044  .046  .030  .323 

S.D.  .050  .068  .044  .355 


0.019 
.032 


TIAL  =  whitebark  pine,  ABLA  =  subalpine  fir,  PIEN 
PSME  =  Douglas-fir. 


Engelmann  spruce. 


and  Engelmann  spruce,  and  greater  than  the  shade- 
intolerant  lodgepole  pine  as  well  (table  1).  Scatterplots  of 
quadrat  densities  vs.  distance  from  the  upper  seed  source 
revealed  interesting  differences  in  regeneration  patterns 
among  the  conifers  (fig.  1).  The  scatterplot  for  whitebark 
pine  was  a  negative  exponential  curve,  with  the  highest 
densities  near  the  seed  source  and  a  long  tail  indicating 


0.512 


0.400  - 


0.200  - 

0 


'3) 
c 

Q 


p.  albicaulis 


•  •    •  •  • 
•  •     •  •  • 

.•    •  .  J  J  I  It         ••  •  • 


0.296  |-      p.  contorta 


0.200 


0.100 


•      •  •  • 

-  •  •      •  • 


•     •  • 


0.08  r       A.  lasiocarpa 


0.05 


0.02 


0.056  r      p  engelmannii 


0.040 


•  •  • 


0.020 


•  I 


•  I  • 
•  •  • 


I  I  »» w    I  l,«  «  i  »»  .  I  »  y  I  «  I  I 


1000 


2000 


3000  3650 


1000 


2000 


3000  3650 


Distance  (m) 


Figure  1 — Scatterplots  of  regeneration  density  per  quadrat  versus  distance  from  the  whitebark  pine 
seed  source  for  four  conifer  species  in  the  Sleeping  Child  Burn.  Clockwise  from  left:  whitebark  pine, 
lodgepole  pine,  Engelmann  spruce,  subalpine  fir.  From  Tomback  and  others  (1990). 


194 


some  regeneration  beyond  3.6  km.  Quadrat  density  vs. 
distance  in  subalpine  fir  showed  a  negative  linear  rela- 
tionship, indicating  that  the  primary  seed  soiirce  for  this 
species  was  also  the  upper,  unbumed  forest.  In  contrast, 
the  scatterplot  for  Engelmann  spruce  was  nearly  flat, 
with  the  highest  values  at  the  lower  end  of  the  transect, 
suggesting  that  seeds  may  be  blown  into  the  ridge  area 
from  both  the  higher  and  lower  seed  sources.  For  lodge- 
pole  pine,  densities  increased  with  distance  from  the  up- 
per, iinbumed  forest  (are  higher  at  lower  elevations). 
Before  the  fire,  the  lower  and  middle  elevations  of  this 
area  were  dominated  by  lodgepole  pine,  and  thus  much  of 
the  postfire  regeneration  may  have  come  from  seeds  from 
serotinous  cones  within  the  bum  (Lotan  1976).  With  the 
exception  of  Engelmann  spruce,  the  density  vs.  distance 
relationships  were  statistically  significant  atP  <  0.01  or 
greater  (Tomback  and  others  1990). 

For  the  Saddle  Moimtain  Burn,  mean  whitebark  pine 
regeneration  density  was  not  significantly  different  than 
that  for  the  Sleeping  Child  Bum,  although  somewhat 
lower  (table  1).  The  density  was  actually  comparable  to 
that  obtained  for  the  ridgetop  quadrats  (west  facing)  in 
the  Sleeping  Child  Bum.  In  contrast,  the  densities  for 
subalpine  fir  and  Engelmann  spruce  were  much  greater 
than  those  in  the  Sleeping  Child  Burn.  Lodgepole  pine 
was  particularly  abimdant  as  well.  Scatterplots  of  quad- 
rat density  vs.  distance  from  the  upper  seed  source  were 
comparable  to  those  from  Sleeping  Child  (fig.  2). 


0.14 


0.10 


r  •• 


(0 

c 

0) 
Q 


0.05 

0 

0.30 
0.20 

0.10 


P.  albicaulis 


•      •  • 


A.  lasiocarpa 


••  • 


•  •  •  • 

•  ■  •  •   •  • 

•  ,  •      ••••••  • 


1000 


2000 


3000 


Distance  (m) 

Figure  2 — Scatterplots  of  regeneration  density 
per  quadrat  versus  distance  from  the  whitebark 
pine  seed  source  for  whitebark  pine  (top)  and 
subalpine  fir  (bottom)  in  the  Saddle  Mountain 
Burn.  From  Tomback  and  others  (1990). 


Table  2 — Maximum  age  differences  within  clusters  of  whitebark 

pine  seedlings.  Cluster  sample  size  for  the  Sleeping  Child 
Burn  =  87;  for  the  Saddle  Mountain  Burn  =  7.  Based  on 
table  4,  Tomback  and  others  (1993) 


Maximum  age 
differences 

Sleeping  Child 

Saddle  Mountain 

Years 

n 

Percent 

n 

Percent 

0 

30 

34.5 

1 

14.3 

1 

24 

27.6 

4 

57.1 

2 

12 

13.8 

1 

14.3 

3 

10 

11.5 

0 

0 

4 

6 

6.9 

1 

14.3 

5 

1 

1.2 

6 

1 

1.2 

7 

2 

2.3 

8 

1 

1.2 

Whitebark  pine  regeneration  showed  a  negative  exponen- 
tial distribution,  and  subalpine  fir  a  similar  but  more  lin- 
ear negative  relationship  between  quadrat  density  and 
distance  from  the  upper,  unbiu-ned  forest  (density  vs. 
distance  relationships  significant  at  P  <  0.001  for  both). 
Similar  to  findings  for  the  Sleeping  Child  Bum,  the  En- 
gelmann spruce  scatterplot  was  nearly  flat.  For  both 
Douglas-fir  and  lodgepole  pine,  density  increased  with 
distance  from  the  upper,  imbumed  forest;  major  seed 
sources  for  both  species  were  at  the  lower  edges  of  the 
bum  and,  in  the  case  of  lodgepole  pine,  seeds  from  within 
the  bum.  However,  the  relationship  between  density  and 
distance  was  not  statistically  significant  for  these  latter 
three  conifers. 

The  distinctive  negative  exponential  shape  of  the  white- 
bark pine  regeneration  curve  probably  results  from  two 
factors:  (1)  whitebark  pine  seeds  are  only  available  at  the 
upper  elevations  of  these  burns,  and  (2)  nutcrackers  tend 
to  bury  higher  densities  of  seeds  near  parent  trees.  Ap- 
parently, nutcrackers  do  cache  seeds  at  long  distances 
from  parent  whitebark  pine  trees,  because  we  found  low 
densities  of  regeneration  up  to  8  km  from  the  whitebark 
pine  seed  source  (Tomback  and  others  1990).  It  could  be 
argued  that  the  concomitant  decrease  in  elevation  in  both 
study  areas  was  a  confounding  variable  in  the  whitebark 
pine  density  vs.  distance  relationship.  However,  recently, 
I  tested  this  possibility  by  examining  whitebark  pine  re- 
generation along  a  3.4-km  transect  following  an  eleva- 
tional  isocline  in  the  Sundance  Bum  of  northern  Idaho. 
Like  the  Sleeping  Child  and  Saddle  Mountain  Bxirns,  the 
whitebark  pine  seed  source  was  restricted  to  one  side  of 
the  bum,  and  the  resulting  regeneration  curve  followed 
a  negative  exponential  distribution  (Tomback  1992). 

It  appears  that  seed  dispersal  by  nutcrackers  has 
provided  whitebark  pine  an  advantage  over  its  shade- 
tolerant  competitor,  subalpine  fir,  in  the  upper  subalpine 
of  the  Sleeping  Child  Burn.  Because  wind-dispersed  fir 
seeds  from  the  upper  seed  source  must  travel  against  the 
prevailing  winds  and  cover  a  long  distance,  the  regenera- 
tion density  of  fir  is  lower  than  that  for  whitebark  pine. 
In  contrast,  in  the  smaller  Saddle  Mountain  Burn,  white- 
bark pine  and  subalpine  fir  have  comparable  regeneration 


195 


densities.  In  this  situation,  dispersal  of  subalpine  fir  seeds 
(and  Engelmann  spruce  seeds)  is  favored  both  by  the  pre- 
vailing winds  from  the  west  and  the  much  smaller  size  of 
the  burn.  Thus,  for  restocking  burns,  seed  dispersal  by 
nutcrackers  may  provide  whitebark  pine  an  advantage 
over  wind-dispersed  conifers  when  (1)  the  size  of  the  burn 
is  large  compared  to  distances  traveled  by  wind-blown 
seeds,  (2)  the  seed  source  is  unfavorably  positioned  with 
respect  to  prevailing  winds,  and  (3)  the  seed  source  is  1  km 
or  more  from  the  perimeter  of  the  burn. 

Another  consequence  of  seed  dispersal  by  nutcrackers 
is  that  seedling  or  sapling  clusters  occurred  at  more  than 
40  percent  of  the  whitebark  pine  regeneration  sites  in  both 
the  Sleeping  Child  and  Saddle  Mountain  Burns  (Tomback 
and  others  1990,  1993).  Cluster  sizes  in  the  Sleeping 
Child  Burn  ranged  from  two  to  eight  stems  with  an  over- 
all mean  of  1.9  stems  per  regeneration  site  (n  =  455  total 
sites).  In  the  Saddle  Mountain  Burn,  clusters  ranged 
from  two  to  10  stems,  with  an  overall  mean  of  2.0  stems 
per  regeneration  site  (n  =  164  sites).  We  aged  two  or  more 
cluster  members  from  a  cluster  subsample  from  each  of 
the  study  areas  (Tomback  and  others  1993).  For  the 
Sleeping  Child  Burn,  there  were  no  differences  in  age 
among  seedlings  within  a  cluster  for  34  percent  of  the 
clusters,  differences  of  1  year  for  28  percent  of  the  clus- 
ters, and  differences  of  2  to  8  years  within  clusters  for  the 
rest  of  the  subsample  (total  n  =  87  clusters;  all  seedlings 
aged  in  63  clusters)  (table  2).  Of  the  small  number  of 
clusters  sampled  in  the  Saddle  Mountain  Burn,  five  clus- 
ters (71  percent)  had  age  differences  of  0  to  1  year,  and 
the  remaining  two  clusters  had  seedlings  differing  by 
2  and  4  years,  respectively  (table  2).  These  resiilts  strongly 
suggest  that  asynchronous  germination  of  seeds  may  oc- 
cur in  some  clusters.  This  could  be  explained  by  delayed 
maturation  of  some  seeds  or  different  sensitivities  to  envi- 
ronmental conditions.  Laboratory  (Leadem  1986;  Pitel  and 
Wang  1990)  and  experimental  field  studies  (McCaughey 
1993)  indicate  that  embryo  underdevelopment  and  de- 
layed germination  are  tjrpical  in  whitebark  pine  seeds. 
Delayed  germination  of  seed  caches  would  result  in  seed- 
ling establishment  following  little  or  no  cone  production 
and,  thus,  continuity  of  regeneration  over  time. 

We  compared  age  structiire  (year  of  tree  establishment 
vs.  time  since  fire)  of  regeneration  for  whitebark  pine 
and  subalpine  fir  in  both  the  Sleeping  Child  and  Saddle 
Mountain  Burns  to  determine  any  important  differences 
(Tomback  and  others  1993).  Both  species  appeared  to  be 
similarly  affected  by  environmental  and  ecological  factors: 
a  disproportionately  high  percentage  of  regeneration  was 
established  between  1977  and  1985,  about  17  to  25  years 
after  fire  (see  fig.  1  in  Tomback  and  others  1993).  Thus, 
major  tree  recruitment  was  episodic.  A  small  increase 
in  mean  January  through  August  precipitation  between 
1977  and  1985  possibly  contributed  to  this  trend.  If  we 
can  generalize  from  these  results,  it  appears  that  white- 
bark pine  and  subalpine  fir  become  established  synchro- 
nously after  fire,  albeit  at  different  densities  depending 
on  relative  seed  dispersal  abilities. 


REGENERATION  IN  THE  GREATER 
YELLOWSTONE  AREA 

Studies  in  the  Greater  Yellowstone  Area  following  the 
major  fires  of  1988  have  provided  comparative  informa- 
tion on  patterns  of  seed  dispersal,  germination,  and  seed- 
ling mortality  in  serai  whitebark  pine  communities 
(Tomback  1991,  1993).  Unlike  the  large-scale,  stand- 
replacing  burns  described  earlier  for  western  Montana, 
the  Yellowstone  fires  created  patches  of  burned  areas  of 
different  severities,  often  surrounded  by  unburned  forest. 

In  1990,  my  research  assistants  and  I  established  a  to- 
tal of  275  permanent  plots,  each  20  m^,  in  two  different 
study  areas.  In  the  Cooke  City  study  area,  Gallatin  Na- 
tional Forest,  we  placed  plots  in  four  sample  sites,  repre- 
senting the  following  treatments:  dry,  severely  biirned 
(50  plots);  moist,  severely  burned  (50  plots);  dry,  imbumed 
(25  plots);  moist,  unburned  (25  plots).  Plot  aspects  in  this 
area  ranged  from  80°  to  180°,  and  elevations  rsmged  from 
2,680  to  2,745  m.  On  Mount  Washbiirn,  Yellowstone 
National  Park,  we  selected  three  sample  sites,  represent- 
ing the  following  treatments:  dry,  severely  burned  (50 
plots);  moist,  severely  burned  (50  plots);  moist,  moder- 
ately burned  (25).  Aspects  ranged  from  250°  to  360°, 
and  elevations  ranged  from  2,560  to  2,745  m.  All  conifer 
regeneration  was  recorded  for  these  plots  in  the  siunmers 
of  1990,  1991,  and  1992;  whitebark  pine  regeneration  was 
mapped  for  each  plot.  I  briefly  summarize  the  results 
here  (Tomback  1991,  1993). 

Whitebark  Regeneration  Trends 

Despite  an  abundant  whitebark  pine  cone  crop  in  1989, 
no  whitebark  pine  regeneration  occurred  on  any  severely 
burned  or  moderately  burned  plots  in  1990  in  either  study 
area  (table  3).  Only  one  plot  in  the  moist,  unburned  treat- 
ment had  one  newly  germinated  seedling  cluster.  Al- 
though there  was  virtually  no  cone  production  in  1990, 
all  severely  burned  and  moderately  burned  treatments 
had  some  new  whitebark  pine  regeneration  in  1991,  which 
was  probably  the  result  of  delayed  germination  of  seeds 
cached  in  1989.  Of  the  unburned  treatments,  again  only 
the  moist  plots  experienced  any  new  regeneration  (table  3). 

In  1992  there  was  some  new  whitebark  pine  regenera- 
tion on  the  Cooke  City  moist,  severely  burned  plots  and 
on  the  Mount  Washburn  moist,  moderately  burned  plots. 
By  1992,  the  latter  sample  site  supported  the  highest  den- 
sity of  whitebark  pine  regeneration  of  all  sites  (table  3). 
Seedling  survivorship  from  1991  to  1992  was  fairly  high. 
Those  that  survived  to  the  summer  of  1992  tended  to  be 
shaded  by  other  plants  or  forest  debris  for  at  least  part 
of  the  day.  One  cluster  of  seedlings  that  germinated  in 
1991  just  outside  a  plot  on  the  Mount  Washburn  moist, 
severely  burned  sample  site  contained  two  new  germinants 
in  1992.  This  observation  confirms  the  possibility  that 
seeds  within  caches  may  germinate  in  different  years,  as 
illustrated  by  the  ages  of  cluster  members  from  western 
Montana. 


196 


Table  3 — Whitebark  pine  regeneration  densities  for  different  plot 

treatments  in  the  Cooke  City  and  Mount  Washburn  study 
areas  following  the  1 988  Yellowstone  fires 


Mean  density  of  regeneration  sites^  per 
Treatment  1990  1991  1992 


Cooke  City 


Dry,  severely  burned 

0 

0.012 

0.015 

Moist,  severely  burned 

0 

.014 

.010 

Dry,  unburned 

0 

0 

.002 

Moist,  unburned 

0.002 

.010 

.008 

Mount  Washburn 

Dry,  severely  burned 

0 

.010 

.011 

Moist,  severely  burned 

0 

.021 

.016 

Moist,  moderately  burned 

0 

.030 

.038 

'A  single  regeneration  site  may  support  either  a  single  seedling  or  a  cluster 
of  seedlings  arising  from  one  seed  cache. 


Regeneration  Trends  of  Other 
Conifers 

Trends  in  the  regeneration  of  other  conifers  differed  be- 
tween the  Cooke  City  and  Mount  Washburn  study  areas, 
probably  as  a  result  of  seed  availability  and  seedbed  con- 
ditions. For  example,  no  conifer  regeneration,  old  or  new, 
occiirred  on  the  Cooke  City  dry,  severely  burned  plots  in 
1990;  small  niombers  of  Engelmann  spruce  and  subalpine 
fir  occurred  on  these  plots  in  subsequent  years.  On  the 
moist,  severely  burned  plots,  new  spruce  germinants 
dominated  in  1990  and  both  spruce  and  fir,  new  and  old, 
in  subsequent  years.  Primarily  new  fir  regeneration  oc- 
curred each  year  on  the  Cooke  City  dry,  unburned  plots, 
with  few  older  firs  surviving;  and,  mostly  spruce  occurred 
on  the  moist,  unburned  plots.  Thus,  spruce  and  fir  regen- 
eration predominated  in  the  Cooke  City  area.  Few  lodge- 
pole  pine  were  apparent  in  the  prefire  forest  or  in  the  un- 
bumed  forest  around  the  treatment  sites. 

In  the  Moimt  Washburn  study  area,  old  and  new  lodge- 
pole  pine  and  new  Engelmann  spruce  occiured  in  1990  on 
the  dry,  severely  biirned  plots.  Older  lodgepole  pine  and 
some  new  spruce  occurred  in  subsequent  years.  On  the 
moist,  severely  burned  plots,  new  lodgepole  pine  and 
spruce  regeneration  predominated  in  1990,  and  older 
lodgepole  pine  and  spruce  in  1991.  In  1992  there  were 
munerous  new  germinants  of  both  species  and  some  sub- 
alpine fir.  High  numbers  of  new  spruce  seedlings  domi- 
nated the  moist,  moderately  burned  plots  in  1990,  and 
older  and  new  spruce  in  subsequent  years.  Thus,  lodge- 
pole pine  and  spruce  regeneration  predominated  in  the 
burned  area  sampled  on  Moimt  Washbiirn.  Lodgepole 
pine  had  been  an  important  species  in  the  prefire  forest 
and  was  present,  along  with  spruce,  in  imburned  forest 
not  far  fi-om  the  plot  treatment  sites.  The  high  spruce  re- 
generation density  on  the  moderately  burned  plots  may 
have  been  the  consequence  of  a  more  favorable  seedbed. 

Although  the  numbers  of  whitebark  pine  germinants 
were  lower  in  most  cases  than  those  of  the  other  conifer 
species  in  both  study  areas,  their  siu-vivorship  from  1991 


to  1992  was  relatively  high.  The  microsites  selected  for 
seed  caching  by  nutcrackers  and  the  hardiness  of  the 
whitebark  pine  seedlings  may  result  in  lower  mortality 
compared  to  other  conifers.  We  must  follow  the  regenera- 
tion patterns  of  whitebark  pine  and  its  forest  associates 
for  several  more  years  to  test  this  idea. 

CONCLUSIONS 

The  patterns  of  whitebark  pine  regeneration  that  we 
observe  are  the  consequence  of  the  seed-dispersal  behav- 
iors and  site  preferences  of  nutcrackers,  as  well  as  their 
tendency  to  bury  multiseed  caches.  Nutcrackers  bury 
seeds  in  burned  terrain  soon  after  fire  (for  example,  see 
Tomback  and  Knowles  1989).  Because  whitebark  pine 
seeds  vary  in  maturation  rate,  nutcrackers  cache  both  de- 
veloped and  imderdeveloped  seeds  (Leadem  1986),  result- 
ing in  germination  over  several  different  years  from  the 
same  seed  crop. 

Early  postfire  regeneration  of  whitebark  pine  may  be  the 
consequence  of  rapid  and  widespread  seed  dispersal  by 
nutcrackers,  suitability  of  the  cache  sites  selected  by  nut- 
crackers, and  the  hardiness  of  the  seedlings  themselves. 

ACKNOWLEDGMENTS 

The  Intermountain  Research  Station,  USDA  Forest  Ser- 
vice, Ogden,  UT,  has  funded  all  whitebark  pine  postfire 
regeneration  studies  in  western  Montana,  northern  Idaho, 
and  the  Greater  Yellowstone  Area.  I  thank  Stephen  Arno, 
Jim  Brown,  and  Bob  Keane  of  the  Intermountain  Re- 
search Station  Fire  Sciences  Laboratory,  Missoula,  MT, 
for  logistic  support  and  many  helpful  discussions  concern- 
ing this  research.  Steve  Arno  and  Bob  Keane  reviewed 
the  draft  manuscript.  Lyn  Hoffinann  and  Sharren  Reuss 
(Sund)  provided  assistance  during  the  western  Montana 
studies,  Kathy  Carsey  and  Mary  Powell  worked  with  me 
during  the  three  field  seasons  of  the  Yellowstone  study, 
and  Mary  Powell,  Jane  Kees,  and  Mark  Flower  were  field 
assistants  for  the  Idaho  work.  Support  for  attending  the 
SSPE  Workshop  was  provided  by  the  College  of  Liberal 
Arts  and  Sciences,  University  of  Colorado  at  Denver. 

REFERENCES 

Arno,  S.  F.  1980.  Forest  fire  history  in  the  northern 
Rockies.  Journal  of  Forestry.  78(8):  460-465. 

Arno,  S.  F.  1986.  Whitebark  pine  cone  crops — a  diminish- 
ing source  of  vdldlife  food?  Western  Journal  of  Applied 
Forestry.  1(3):  92-94. 

Arno,  S.  F.;  Petersen,  T.  D.  1983.  Variation  in  estimates 
of  fire  intervals:  a  closer  look  at  fire  history  on  the  Bit- 
terroot  National  Forest.  Res.  Pap.  INT-301.  Ogden,  UT: 
U.S.  Department  of  Agricult;ire,  Forest  Service,  Inter- 
mountain Forest  and  Range  Experiment  Station.  8  p. 

Arno,  S.  F.;  Hofif,  R.  J.  1989.  Silvics  of  whitebark  pine 
{Pinus  albicaulis).  Gen.  Tech.  Rep.  INT-253.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermountain  Research  Station,  lip. 


197 


Finklin,  A.  I.  1983.  Weather  and  climate  of  the  Selway- 
Bitterroot  Wilderness.  Moscow,  ID:  University  Press 
of  Idaho.  144  p. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Leadem,  C.  L.  1986.  Seed  dormancy  in  three  Pinus  species 
of  the  Inland  Mountain  West.  In:  Shearer,  R.  C,  comp. 
Proceedings — conifer  tree  seed  in  the  Inland  Mountain 
West  symposium;  1985  August  5-6;  Missoula,  MT.  Gen. 
Tech.  Rep.  INT-203.  Ogden,  UT:  U.S.  Department  of 
Agrictdture,  Forest  Service,  Intermountain  Research 
Station:  117-124. 

Lotan,  J.  E.  1976.  Cone  serotiny — fire  relationships  in 
lodgepole  pine.  In:  Lyon,  L.  J.,  comp.  Tall  Timbers  Fire 
Ecology  Conference  No.  14  and  Intermountain  Fire  Re- 
search Council  fire  and  land  management  symposium: 
proceedings;  Missoula,  MT:  267-278. 

McCaughey,  W.  W.;  Schmidt,  W.  C;  Shearer,  R.  C.  1986. 
Seed  dispersal  characteristics  of  conifers  in  the  Inland 
Mountain  West.  In:  Shearer,  R.  C,  comp.  Proceedings — 
conifer  tree  seed  in  the  Inland  Mountain  West  sympo- 
sium; 1985  August  5-6;  Missoula,  MT.  Gen.  Tech.  Rep. 
INT-203.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Research  Station:  50-62. 

McCaughey,  W.  W.  1993.  Delayed  germination  and  seed- 
ling emergence  of  Pinus  albicaulis  in  a  high  elevation 
clearcut  in  Montana,  USA.  In:  Edwards,  D.  G.,  comp. 
Seed  dormancy  and  barriers  to  germination:  proceed- 
ings lUFRO  symposium;  1991  April  23-26;  Victoria,  BC. 
Victoria,  BC:  Forestry  Canada.  [In  press]. 

Pfister,  R.  D.;  Kovalchik,  B.  L.;  Arno,  S.  F.;  Presby,  R.  C. 
1977.  Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  174  p. 

Pitel,  J.  A.;  Wang,  B.  S.  P.  1990.  Physical  and  chemical 
treatments  to  improve  germination  of  whitebark  pine 
seeds.  In:  Schmidt,  W.  C;  McDonald,  K.  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
130-133. 

Romme,  W.  H.  1980.  Fire  frequency  in  subalpine  forests 
of  Yellowstone  National  Park.  In:  Fire  history  work- 
shop: proceedings;  1984  October  20-24;  Tucson,  AZ. 


Gen.  Tech.  Rep.  RM-81.  Fort  Collins,  CO:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Rocky  Mountain 
Forest  and  Range  Experiment  Station:  27-30. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. The  Living  Bird.  16:  123-161. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.  1991.  Progress  report  on  1991  fieldwork 
to  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station,  Ogden,  UT:  patterns  of 
postfire  regeneration  of  whitebark  pine  in  the  Greater 
Yellowstone  Area. 

Tomback,  D.  F.  1992.  Patterns  of  postfire  whitebark  pine 
regeneration  in  the  Sundance  Burn  of  northern  Idaho. 
Unpublished  data  on  file  at:  Department  of  Biology, 
University  of  Colorado  at  Denver,  Denver,  CO. 

Tomback,  D.  F.  1993.  Progress  report  on  1992  fieldwork 
to  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station,  Ogden,  UT:  patterns  of 
postfire  regeneration  of  whitebark  pine  in  the  Greater 
Yellowstone  Area. 

Tomback,  D.  F.;  Knowles,  J.  W.  1989.  Observations  on 
nutcracker  seed  harvesting  and  caching  on  Mount 
Washburn  following  the  1988  Yellowstone  fires.  Unpub- 
lished data  on  file  at:  Department  of  Biology,  University 
of  Colorado  at  Denver,  Denver,  CO. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Co- 
evolution  of  nutcrackers  and  pines:  implications  for 
forest  regeneration.  In:  Schmidt,  W.  C;  McDonald, 
K.  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  118-129. 

Tomback,  D.  F.;  Sund,  S.  K.;  Hoffmann,  L.  A.  1993.  Post- 
fire regeneration  of  Pinus  albicaulis:  height-age  rela- 
tionships, age  structure,  and  microsite  characteristics. 
Canadian  Journal  of  Forest  Research.  23:  113-119. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of 
the  Clark's  nutcracker  and  the  pinon  pine  for  efficient 
seed  harvest  and  dispersal.  Ecological  Monographs.  47: 
89-111. 


198 


NEW  TREND  IN  DENDROCHRONOLOGY: 
1.  THEORETICAL  PRINCIPLES  OF 
REPROCHRONOLOGY 

Vladislav  N.  Vorobjev 


Abstract — Reviews  previous  research  pertinent  to  tree  growth 
and  cone-bearing  relationships  in  Siberian  stone  pine  (Pinus 
sibirica)  and  other  species.  The  ultimate  objective  is  to  secure 
stable  and  moderately  high  cone  production  without  significantly 
jeopardizing  tree  growth.  To  reach  this  objective  the  author 
points  out  potential  avenues  of  research  in  tree  growth  dynamics 
and  reproductive  activities,  utilizing  procedures  used  in 
dendrochronology. 


Dendrochronology  was  established  as  a  science  of  the 
growth  dynamics  of  trees,  the  structure  of  the  annual 
ring,  and  the  effect  of  internal  and,  especially,  external 
factors.  The  latter  determined  the  development  of  den- 
drochronology and  a  number  of  other  special  scientific 
areas. 

Generative  processes  have  been  studied  separately  and 
dealt  with  by  the  corresponding  branch  of  seed  farming  in 
terms  of  the  concept  of  the  crop  dynamics  or  crop  cyclicity. 
In  doing  so,  novel  independent  results  have  been  obtained 
and,  more  important,  premises  have  been  created  for  com- 
bining these  findings  with  dendrochronology. 

The  correlation  between  the  growth  and  generative  de- 
velopment of  higher  plants  comprises  a  number  of  aspects 
and  has  been  studied  for  a  long  time.  The  study  of  herba- 
ceous plants  has  lagged  behind  that  of  tree  investigations. 
Suffice  it  to  say  that  not  only  have  monographs  on  this 
subject  been  lacking,  but  original  publications  contribut- 
ing to  this  field  have  also  been  scarce.  The  only  exception, 
to  the  best  of  oiu-  knowledge,  has  been  the  work  done 
in  Tomsk  by  the  laboratory  of  cone-bearing  trees 
(Goroshkevich  1989;  Nekrasova  1974;  Vorobjev  1983; 
Vorobjev  and  Vorobjeva  1982;  Vorobjev  and  others  1989). 
However,  the  problem  of  the  relation  between  growth  and 
cone  bearing  (in  coniferous  trees,  in  particular)  is  of  great 
importance  in  view  of  the  necessity  to  increase  cone  crops. 
While  this  problem  is  being  solved  in  herbaceous  plants 
and  fi-uit-bearing  trees  by  an  intense  short-term  shift  of 
metabolism  toward  activization  of  the  generative  pro- 
cesses, this  technique  fails  in  coniferous  trees  because 
of  the  long  growth  period  inherent  in  them. 

It  is  assumed  on  the  basis  of  the  foregoing  consider- 
ations that  control  of  the  sexual  reproduction  of  trees 


Poster  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Vladislav  N.  Vorobjev  is  Head  of  the  Academy  of  Sciences,  Siberian 
Branch,  Institute  for  Ecology  of  Natural  Complexes,  Academichesky  pr.  2, 
Academgorodok,  Tomsk,  634055,  Russia. 


must  be  related  to  the  determination  of  optimal  correla- 
tions between  the  generative  and  growth  processes 
through  the  long-term  activization  of  the  vital  activity 
of  plants. 

PREVIOUS  RESULTS 

The  expedience  of  this  approach  was  also  confirmed 
by  the  results  of  our  previous  investigations.  The  experi- 
ments on  the  resin  tapping  of  the  Siberian  cedar  pine, 
sometimes  called  Siberian  stone  pine  {Pinus  sibirica), 
showed  that  the  increase  of  cone  crops  by  creating  favor- 
able conditions  for  cone  bearing  alone  had  no  futiu'e  be- 
cause it  would  eventually  resiilt  in  the  weakening  of 
the  growth  processes  and  then  of  the  generative  ones 
(Vorobjev  1974).  Therefore,  it  appears  necessary  to  look 
for  certain  correlations  between  growth  and  cone  bearing 
that  would  make  it  possible  to  secure  moderately  high 
cone  crops  on  a  long-term  stable  basis.  It  is  this  problem 
that  seems  to  be  of  great  theoretical  and  practical  interest 
(Vorobjev  and  Vorobjeva  1980). 

The  study  of  the  effect  of  the  reproductive  activity  on 
shoot  growth  and  the  shoot  condition  on  subsequent  cone 
bearing  provides  insight  into  the  relation  between  tree 
growth  and  cone  bearing.  Observations  have  shown  that 
cone  bearing  depends  on  the  previous  shoot  growth  in 
Scotch  pine  (Pinus  sylvestris),  English  oak  (Quercus  sp.), 
and  apple  (Malus  sp.)  trees  (Minina  1954;  Polozova  1957; 
Pravdin  1950;  Tsel'niker  and  Semikhatova  1957).  It  has 
been  pointed  out  that  optimal  shoot  growth  is  required  to 
induce  cone  bearing.  Of  no  less  importance  is  the  time  it 
takes  to  achieve  growth,  since  the  flower  germs  are  initi- 
ated when  apical  growth  gets  weaker  or  terminates.  The 
surface  area  of  leaves  also  plays  a  significant  role. 

The  effect  of  cone  bearing  on  the  current  increment  has 
been  the  subject  of  a  number  of  papers.  In  particular,  the 
diameter  increment  in  coniferous  trees  was  shown  to  de- 
crease in  the  cone-maturing  year  (Chalupka  and  others 
1975, 1976, 1977;  Danilov  1953).  However,  the  effect  of  the 
decreased  diameter  increment  on  fiirther  seminiference  of 
trees  and  metamere  growth  has  not  been  examined  yet. 

The  change  in  shoot  growth  due  to  cone  bearing  is  the 
net  effect  of  sexualization  and  subsequent  development 
of  generative  organs  at  different  stages  of  their  maturity. 
The  dependence  of  the  cambial  growth  of  the  shoot  on  its 
sex  has  been  described  in  literature  (Nekrasova  1972; 
Varnell  1976).  However,  no  experimental  evidence  of  the 
role  of  sexualization  and  nutrition  of  the  generative  or- 
gans, and  of  the  effect  of  the  nutrition  of  the  1-  or  2-year 
cones  on  the  annual  ring  and  its  internal  structure  in  the 
growing  shoot  has  been  available  (Vorobjev  and  Vorobjeva 
1982). 


199 


RETROSPECTIVE  METHOD 

Good  possibilities  for  the  study  of  the  effect  of  cone 
crops  on  tree  increment  (both  Hnear  and  cambial)  have 
opened  up  due  to  recent  development  of  the  method  of  ret- 
rospective accoimt  of  the  reproductive  activity  of  trees 
during  their  lifetime  (see  Vorobjev  and  others,  these  pro- 
ceedings). Significantly,  this  procedure  can  also  be  used 
to  reproduce  male  "flowering"  and  the  formation  of  a  num- 
ber of  vegetative  metameres,  though  for  a  shorter  period 
of  time.  In  this  case,  shoots,  branches,  and  crown  become 
major  objects  of  inquiry  in  the  investigation  of  the  relation 
between  the  growth  and  generative  development  of  trees. 

The  research  done  into  the  growth-cone-bearing  rela- 
tionship in  Siberian  cedar  pine  in  West  Siberia  and  Gomy 
Altai  for  the  last  100  years  or  more  (Vorobjev  1983)  sug- 
gests a  new  trend  in  the  investigations  made  at  the  cross- 
roads of  the  two  sciences,  biology  of  reproductive  develop- 
ment and  dendrochronology. 

This  trend  can  be  referred  to  as  reprochronology,  com- 
prising both  the  cyclic  fluctuations  of  the  reproductive  ac- 
tivity and  its  relation  to  the  cambial  growth.  Perhaps,  the 
word  is  not  a  happy  choice  from  the  terminology  point  of 
view,  but  it  generally  implies  sexual  reproduction. 

The  time  series  of  reproductive  activity  studied  show 
cycles  of  variable  duration  ranging  from  3  to  28  years 
or  more.  A  10-  to  11-year  reproductive  cycle  is  less  pro- 
noimced  than  the  growth  cycle.  The  relation  between  the 
crop  dynamics  and  solar  activity  is  feebly  marked,  and 
seed  cycles  fall  both  at  the  peak  and  decay  of  the  solar 
activity.  The  reproductive  activity  is  indirectly  related  to 
the  cambial  growth  and  is  delayed  with  respect  to  the  lat- 
ter by  the  time  interval  determined  by  the  life  cycle  of  one 
generation  and  the  growth  aftereffect.  The  peak  of  the 
siminiference  activity  generally  falls  at  the  descending 
branch  of  the  solar  cycle  and  the  period  of  the  retarded 
shoot  growth.  An  increase  in  cone  crops  is  related  to  the 
previous  solar  activity  and  vigorous  growth. 

Thus  it  appears  worthwhile  to  study  further  the  struc- 
ture of  the  time  series  of  the  reproductive  activity  and  their 
relation  to  the  dynamics  of  growth  of  shoots,  branches,  and 
trees,  to  evaluate  the  effect  of  the  external  factors  using 
new  methods,  and  to  develop  prediction  models.  Also  of 
scientific  importance  are  individual,  interpopulational, 
geographical,  ecological,  and  other  aspects  of  the  relation 
between  the  growth  and  the  generative  development  of 
trees  as  well  as  between  the  seminiference  and  the  annual 
ring  of  the  shoot  and  tree.  It  should  be  noted  that  the 
foregoing  potential  research  areas  suggest  the  use  of  the 
techniques  and  procedures  adopted  in  dendrochronology. 

REFERENCES 

Chalupka,  W.;  Giertych,  M.;  Krolikowski,  Z.  1975.  The 
effect  of  cone  crops  on  growth  in  Norway  spruce  {Picea 
abies  (L.)  Karst.).  Arbor.  Kor.  20:  201-212. 


Chalupka,  W.;  Giertych,  M.;  Krolikowski,  Z.  1976.  The 
effect  of  cone  crops  in  Scotch  pine  on  tree  diameter  in- 
crement. Arbor.  Kor.  21:  361-366. 

Chalupka,  W.;  Giertych,  M.;  Krolikowski,  Z.  1977.  Rela- 
tion between  specific  gravity  of  wood  in  Norway  spruce 
(Picea  abies  (L.)  Karst.):  some  growth  parameters  and 
cone  yield.  Arbor.  Kor.  22:  205-212. 

Danilov,  D.  N.  1953.  Effect  of  cone-bearing  on  tree-ring 
structure  in  common  spruce  (Picea  excelsa).  Bot.  J. 
38(3):  367-377.  [In  Russian]. 

Goroshkevich,  S.  N.  1989.  The  dynamics  of  the  micro- 
strobile  initiation  in  the  male  shoots  of  the  Siberian  ce- 
dar pine  as  related  to  weather  conditions.  Ecol.  6:  33-39. 
[In  Russian]. 

Minina,  E.  G.  1954.  Biological  principles  of  flowering  and 
fruiting  oak.  Transactions  of  Inst,  of  Forest  of  USSR 
Acad,  of  Sci.  17:  5-97.  [In  Russian]. 

Nekrasova,  T.  P.  1972.  Biological  principles  of  semi- 
niference of  Siberian  cedar  pine.  Novosibirsk:  Nauka. 
272  p.  [In  Russian]. 

Nekrasova,  T.  P.  1974.  The  cyclicity  of  the  cone-bearing  of 
Siberian  cedar  pine.  In:  Biology  of  seed  reproduction  of 
coniferous  trees  of  West  Siberia.  Novosibirsk:  Nauka: 
70-75.  [In  Russian]. 

Polozova,  (L.)  Y.  1957.  Growth  of  oak  shoots  as  related  to 
fruiting  processes.  Letters  of  Inst,  of  Forest  of  USSR 
Acad,  of  Sci.  9:  9-17.  [In  Russian]. 

Pravdin,  (L.)  F.  1950.  Sexual  dimorphism  in  Scotch  pine 
(Pinus  sylvestris  L.).  Transactions  of  Inst,  of  Forest  of 
USSR  Acad,  of  Sci.  3:  190-201.  [In  Russian]. 

Tsel'niker,  Y.  L.;  Semikhatova,  O.  A.  1957.  On  the  rela- 
tion between  vegetative  and  generative  stages  of  shoot 
development  in  certain  trees.  Bot.  J.  42(7):  1044-1054. 
[In  Russian]. 

Vamell,  R.  J.  1976.  Correlation  between  branch  charac- 
teristics and  production  of  female  strobile  in  longleaf 
pine.  Forest  Science.  22(2):  159-161. 

Vorobjev,  V.  N.  1974.  Growth  and  generative  processes 
in  Siberian  cedar  pine  under  metabolic  disturbance.  In: 
Biology  of  seed  reproduction  of  coniferous  trees  of  West 
Siberia.  Novosibirsk:  Nauka:  140-170.  [In  Russian]. 

Vorobjev,  V.  N.  1983.  Biological  principles  of  compre- 
hensive utilization  of  Siberian  cedar  pine  forests. 
Novosibirsk:  Nauka.  254  p.  [In  Russian]. 

Vorobjev,  V.  N.;  Vorobjeva,  N.  A.  1980.  On  the  problem  of 
regulation  of  seed  crops  of  coniferous  trees.  In:  Selec- 
tion, genetics  and  seed  farming  of  trees  as  the  basis  for 
forming  of  highly  productive  forests.  Moscow:  391-395. 
[In  Russian]. 

Vorobjev,  V.  N.;  Vorobjeva,  N.  A.;  Goroshkevich,  S.  N. 
1989.  Growth  and  sex  of  Siberian  cedar  pine. 
Novosibirsk:  Nauka.  167  p.  [In  Russian). 

Vorobjeva,  N.  A.;  Vorobjev,  V.  N.  1982.  The  effect  of  fe- 
male generative  organs  on  the  shoot  growth  in  the 
Siberian  cedar  pine.  Lesovedeneye  (For.  Sci.)  1:  51-57. 
[In  Russian]. 


200 


NEW  TREND  IN  DENDROCHRONOLOGY: 
2.  METHOD  OF  RETROSPECTIVE 
STUDY  OF  SEMINIFERENCE  DYNAMICS 
IN  PINACEAE 

V.  N.  Vorobjev 

S.  N.  Goroshkevich 

D.  A.  Savchuk 


Abstract — Introduces  a  new  method  of  retrospective  analysis  of 
cone  crop  dynamics  of  Siberian  stone  pine  {Pinus  sihirica)  for  up 
to  100  years  before  the  present.  For  about  the  first  20  years,  cone 
abscissions  are  discernible  on  the  exterior  of  the  shoots  and  can 
be  used  to  reconstruct  cone  crop  production,  but  after  this  they 
are  not  identifiable.  A  method  was  developed  utilizing  cross- 
sectional  saw  cuts  to  examine  the  interior  of  the  branch  shoots. 
Traces  of  the  abscissions  of  cones  remain  in  the  shoots  and,  con- 
sequently, cone  crops  can  be  reconstructed  for  virtually  the  entire 
life  of  the  tree  with  the  procedures  described  in  this  paper.  This 
method  may  also  be  useful  for  other  conifers. 


A  retrospective  study  of  the  life  activity  of  trees  has  al- 
ways been  of  great  interest  to  scientists.  Retrospective 
analysis  has  been  widely  practiced  in  dendrochronologic 
and  dendroclimatic  studies  focusing  on  the  investigation 
of  tree  rings  whose  structure  and  dimensions  are  impor- 
tant indicators  of  the  life  activity  of  the  tree  organism. 
The  variations  of  tree  increment  observed  are  basically 
related  to  the  effect  of  exogenic  factors,  and  cHmatic  con- 
ditions in  particular.  As  to  the  endogenic  factors  and,  pri- 
marily, the  reproductive  activity,  their  contribution  is  rec- 
ognized but  not  accounted  for  because  of  the  difficulty  of 
estimating  cone  crops  for  long  periods  of  time.  In  fact, 
this  aspect  of  activity  of  the  tree  organism  is  regarded 
as  the  so-called  "white  noise"  in  a  multifactor  dendrochro- 
nological  analysis  of  the  effect  of  different  conditions. 

The  early  information  on  the  aimual  dynamics  of  cone 
crops  has  been  obtained  from  direct  counts  of  the  number 
of  generative  organs  for  several  years.  This  technique  is 
fairly  widely  used  even  today.  However,  applications  of 
this  technique  are  Hmited  due  to  the  necessity  of  long- 
term  observations.  Naturally,  scientists  sought  to  develop 
a  method  of  retrospective  study  of  the  cone  crop  dynamics. 
This  method  is  based  on  the  fact  that  the  cones  and  fruits 
either  remain  on  the  branches  after  the  seed  abscission 
or  fall  out,  leaving  traces  on  the  shoot  bark.  The  seasonal 


Poster  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

V.  N.  Vorobjev  is  Head  of  and  S.  N.  Goroshkevich  and  D.  A.  Savchuk  are 
Members  of  the  Academy  of  Sciences,  Siberian  Branch,  Institute  for  Ecol- 
ogy of  Natural  Complexes,  2,  Acadcemichesky  pr.,  Academgorodok,  Tomsk, 
634055,  Russia. 


character  of  the  shoot  growth  in  trees  in  the  temperate 
zone  makes  it  possible  to  age  each  shoot  and,  conse- 
quently, its  generative  organs. 

The  first  information  we  are  aware  of  on  the  tise  of  the 
retrospective  method  of  studying  the  crop  dynamics  can 
be  foimd  in  the  work  of  A.  Renvall  (1912)  for  the  Scotch 
pine  and  N.  Nesterov  (1914)  for  some  leaf -bearing  trees 
(oak,  maple,  hawthorn).  The  method  was  further  devel- 
oped and  elaborated  in  great  detail  by  Z.  Trofimova 
(1953),  T.  Nekrasova  (1957),  and  P.  Gorchakovsky  (1958) 
for  different  coniferous  trees.  A  most  comprehensive  de- 
scription of  this  method  was  given  by  A.  Korchagin  (1960). 

The  advent  of  the  retrospective  method  has  facilitated 
significant  progress  in  the  investigation  of  crop  dynamics. 
The  latter  can  be  reproduced  for  the  past  10-15  years  on 
the  basis  of  the  cone  traces  on  the  shoot  bark.  Special  em- 
phasis has  been  placed  on  Siberian  cedar  pine  (sometimes 
called  Siberian  stone  pine)  because  of  the  great  natural 
and  economic  importance  of  its  seed  crops  (Iroshnikov 
1963;  Vorobjev  1964,  1974).  As  a  result,  a  cycUc  charac- 
ter of  the  fluctuations  in  the  crops  for  different  years  was 
revealed  and  the  relations  between  the  seminiference, 
weather  conditions,  and  tree  grovTth  were  examined. 
However,  further  advance  of  the  investigations  required 
that  the  methodology  be  improved. 

In  our  opinion,  there  are  two  very  essential  ways  of  im- 
proving the  method.  First,  work  out  a  special  technique 
for  reconstructing  the  stages  of  the  development  of  each 
generation  from  the  initiation  to  the  complete  mattiration 
of  cones  and  to  determine  the  number  of  cones  that  pre- 
maturely abscissed  in  different  stages  of  their  develop- 
ment. These  data  will  make  it  possible  to  study  the  crop 
dynamics  and,  in  particular,  its  effect  on  the  tree  and 
shoot  growth  and  the  annual  ring  structure.  Second,  ex- 
tend the  period  of  the  retrospective  account  of  crops  from 
10-15  years  to  100  years  or  more. 

The  object  of  the  present  paper  is  to  introduce  a  new 
method  of  retrospective  analysis  of  the  seminiference  dy- 
namics of  Siberian  cedar  pine,  making  use  of  the  forego- 
ing information. 

CONE  DEVELOPMENT 

Cones  in  the  Siberian  cedar  pine  are  initiated  in  late 
July  or  early  August  (Nekrasova  1972).  The  germs  of  the 
female  cones  hibernate  in  the  form  of  undifferentiated 


201 


Figure  1 — Female  cone  traces  on  the  shoot  bark,  x  6: 
(a)  cone  that  abscissed  before  pollination,  (b)  cone  that 
abscissed  after  pollination,  (c)  cone  that  abscissed  during 
maturation,  (d)  mature  cone. 


primordia  surrounded  by  bud  scales.  The  primordia  are 
located  in  the  distal  part  of  the  winter  terminal  bud.  The 
development  of  the  female  cones  begins  in  May.  An  active 
linear  growth  of  shoots  occurs  in  the  first  half  of  June, 
and  the  cones  resemble  large  lateral  buds  in  the  distal 
part  of  the  growing  shoot.  Special  observations  show  that 
the  development  of  some  of  the  female  cones  ceases  at  this 
period  of  time  and  they  abort.  For  example,  normally  de- 
veloping cones  were  bright  green  and  10-15  mm  long  in  a 
Siberian  cedar  pine  forest  near  a  settlement  in  the  south 
of  the  Tomsk  region  on  J\me  6,  1990.  Some  of  the  female 
buds  began  to  burst,  while  others  were  of  much  smaller 
size.  In  the  latter  case,  a  longitudinal  section  of  a  germ 
showed  that  it  had  died  in  the  differentiation  phase.  It 
looked  like  a  small  (1.5-2.0  mm)  brown  protuberance. 
After  the  degeneration  of  the  germinal  cone  the  bud  dries 
up  and  abscisses.  Abortive  female  buds  leave  small  (2.5- 
3.0  mm)  dark  triangular  traces  (fig.  la). 

Pollination  in  Siberian  cedar  pine  generally  occurs  in 
the  second  half  of  July  (Nekrasova  1972).  After  pollina- 
tion the  cones  grow  for  some  time  until  they  are  2  mm 
long.  During  this  period  (June-July)  the  cones  may  also 
cease  their  development  and  abort.  Traces  of  cones  that 
abscissed  shortly  after  the  pollination  are  larger  (4-5  mm), 
and  more  roundish  than  those  of  the  abortive  female  buds 
and  are  relatively  smooth  surfaced  (fig.  lb). 


In  the  year  of  maturation  both  the  cones  and  shoots 
grow  larger.  The  cones  reach  their  ultimate  dimensions 
by  the  time  of  fertilization  (Jvme  15-20)  (Nekrasova  1972). 
Therefore,  the  cones  that  absciss  in  the  process  of  matura- 
tion (June-August)  leave  the  traces  of  the  same  shape  and 
size  as  normal  mature  cones  do.  The  traces  are  large  (up 
to  7-8  mm)  and  oval  with  a  funnel-shaped  inner  surface 
(fig.  Ic,  d).  Mature  cones  absciss  in  late  August  or  early 
September.  The  inner  surface  of  their  traces  on  the  shoot 
bark  is  covered  with  resin  (fig.  Id). 

TRACE  IDENTIFICATION 

The  identification  of  cone  traces  would  require  knowl- 
edge of  the  morphostructure  of  the  mature  shoot  and  the 
origin  of  its  different  parts.  The  annual  shoot  in  large 
cone-bearing  branches  in  Siberian  cedar  pine  consists  of 
two  elementary  shoots  (Vorobjev  and  others  1989).  The 
first  (spring)  shoot  develops  from  the  winter  bud  in  June. 
The  second  (summer)  shoot  is  initiated  in  June  and  is 
found  to  enlarge  at  the  beginning  of  Jvdy.  Each  elemen- 
tary shoot  has  short  shoots  (brachyblasts)  and  lateral  ex- 
tending shoots.  The  length  of  the  axis  and  the  nvunber  of 
its  metameres  in  the  spring  shoot  are  approximately  ten 
times  as  large  as  in  the  summer  shoot.  The  female  cones 
are  located  in  the  distal  part  of  the  spring  shoot.  Small 
lateral  vegetative  shoots  usually  grow  at  the  same  level 


202 


as  the  female  cones.  However,  there  may  be  no  lateral 
shoots.  The  main  whorl  of  large  lateral  shoots  is  located 
on  the  simimer  shoot.  It  shoiild  be  noted  that  simimer 
shoots  in  mature  trees  are  formed  when  the  spring  shoot 
bears  cones.  Therefore,  the  summer  shoot  can  be  regarded 
as  an  indirect  evidence  of  the  presence  of  cones  in  the 
spring  shoot.  The  traces  of  mature  cones  on  mature  need- 
less shoots  will  occur  in  the  distal  part  of  each  annual 
shoot  under  the  main  whorl  of  the  lateral  shoot. 

The  traces  should  be  counted  from  the  top  of  the  branch 
to  yield  the  number  of  different  types  of  traces  on  each 
annual  shoot.  This  procediire  makes  it  possible  to  recon- 
struct the  entire  dynamic  cycle  of  the  initiation,  develop- 
ment, and  matiiration  of  cones  (over  10  to  12  years).  The 
traces  of  abortive  female  buds  on  older  shoots  are  de- 
stroyed because  of  the  secondary  bulge  of  the  shoot  axis 
and  core  cracking,  which  precludes  their  identification 
and  count.  Large  traces  of  mature  cones  persist  for  a 
longer  period  of  time,  and  they  can  be  used  for  the  recon- 
struction of  crop  dynamics  over  approximately  20  years. 

SAMPLING 

Our  experience  shows  that  it  is  necessary  to  sample 
10  branches  from  one  tree.  The  branches  selected  must 
be  from  different  parts  of  the  crown  (top,  middle,  lower). 
Each  branch  must  be  no  less  than  20  years  old.  To  recon- 
struct the  djTiamics  of  the  initiation  and  development  of 
cones  on  individual  branches,  one  has  to  use  branches  of 
mediimi  size  and  ignore  the  traces  of  cones  on  the  lateral 
shoots.  On  the  other  hand,  the  seminiference  dynamics 
for  the  tree  or  stand  as  a  whole  can  be  followed  by  using 
cone-bearing  branches  of  different  size  and,  most  impor- 
tant, by  taking  into  account  the  traces  on  small  lateral 
branches. 

As  the  number  of  annual  rings  increases,  the  traces 
of  cones  on  the  bark  in  the  cone-bearing  branches  of  trees 
become  indiscernible.  Therefore,  a  procedure  was  devel- 
oped for  the  identification  of  such  traces  on  the  cross- 
sectional  saw  cuts  of  branches  (Vorobjev  1979).  The  saw 
cuts  of  the  bases  of  whorls  show  that  the  traces  of  cones 
remain  within  shoots  and  are  quite  discernible,  which 
makes  it  possible  to  reconstruct  the  crop  dynamics  virtu- 
ally for  the  entire  Ufe  of  the  tree.  Thus,  the  time  scale  of 
the  proposed  procedure  can  be  extended. 

The  saw  cuts  must  be  made  1  cm  beneath  the  point 
where  the  longitudinal  axis  of  the  branch  intersects  the 
axes  of  the  lateral  shoots.  Then  they  are  paper-sanded 
and  examined  with  a  stereomicroscope  x  16.  The  saw  cuts 
must  begin  with  the  whorls  where  the  external  traces  on 
the  bark  are  fairly  visible.  The  identification  of  the  inner 
traces  should  also  begin  with  those  whorls.  This  is  espe- 
cially important  for  the  differentiation  of  different  periods 
of  cone  development. 

TRACE  CHARACTERISTICS 

The  cone  traces  on  the  saw  cuts  of  branches  can  be 
described  as  follows:  The  trace  of  the  cone  that  abscissed 
before  pollination  (the  trace  of  the  abortive  female  bud) 
has  the  form  of  a  brown  ray  with  a  uniform  width  of 


0.3±0.03  mm  and  a  length  of  1.6±0.2  mm.  It  runs  from 
the  pith  of  the  shoot  to  its  bark  at  the  backgrotmd  of  a 
yellow  saw  cut  wood  (fig.  2a).  The  trace  of  the  cone  that 
abscissed  after  pollination  and  that  of  the  immature  cone 
are  half  as  large  edgewise  as  the  trace  of  the  cone  that 
abscissed  before  pollination.  The  latter  trace  (fig.  2b)  in 
contrast  to  the  former  (fig.  2c)  has  a  bulge  at  the  end  (1.7 
times  as  large  as  the  ray  width)  or  is  slightly  forked.  The 
trace  of  the  mature  cone  (fig.  2d)  is  twice  as  large  as  that 
of  the  immature  cone.  Note  that  there  may  be  no  ray. 
The  trace  of  the  mature  cone  differs  greatly  from  all  the 
other  traces  (see  fig.  2d  and  figs.  2a,  b,  c)  by  an  oval  or 
triangular  bulge  (2.4+0.3  mm  wide)  at  the  ray  end.  Such 
traces  are  often  accompanied  by  a  specific  deformation  of 
the  tree  rings  as  a  result  of  rupture  of  phloem  and  xylem 
tissues  of  the  shoot  by  a  great  n;unber  of  cones.  Thus  the 
more  mature  the  cone  the  brighter  and  longer  is  its  trace 
on  the  saw  cut.  The  bulge  is  also  larger. 

The  traces  of  cones  differ  from  the  traces  of  the  lateral 
shoots  and  brachyblasts  on  the  cross-sectional  saw  cuts. 
The  trace  of  the  lateral  shoot  (fig.  2e)  is  wide  and  gener- 
ally runs  toward  the  bark  to  join  the  shoot  axis.  The  trace 
of  the  cone  does  not  reach  the  bark.  The  trace  of  brachy- 
blast  (fig.  2f)  is  very  similar  to  the  trace  of  the  cone  that 
abscissed  before  pollination  (the  abortive  female  bud),  but 
the  former  trace  is  much  longer  (2.5+0.2  mm)  and  its  end 
grows  narrower.  The  main  thing  is  that  the  trace  of 
brachyblast  is  surrounded  by  a  brighter  layer  of  wood 
(3.8±0.3  X  0.7±0.1  mm),  which  is  visible  to  a  naked  eye. 

USE  OF  METHODS 

We  feel  that  use  can  be  made  of  methods  based  on  the 
difference  in  the  density  of  the  cone  traces  and  that  of  the 
surrounding  wood  (roentgenography,  densitography,  laser 
techniques,  etc.)  in  the  cases  where  the  traces  on  the  saw 
cuts  are  dim  (especially  the  traces  of  the  prematvirely 
abscissed  cones),  or  else,  if  the  saw  cut  has  been  made  in 
a  wrong  way.  In  particular,  soft  x-ray  roentgenography 
of  a  5-  to  7-mm-thick  saw  cut  of  a  branch  (the  time  of  expo- 
sure was  6  minutes)  made  it  possible  to  determine  unam- 
biguously the  structure  of  tree  rings,  cone  traces,  and  those 
of  other  metameres  (Vorobjev  1979).  The  above  techniques 
provide  docvmientary  evidence  for  subsequent  analysis. 

This  ability  to  distinguish  the  internal  traces  must  be 
practiced  using  the  branches  with  fairly  visible  external 
cone  traces  on  the  bark.  The  selection  of  the  branches 
for  the  analysis  must  be  made  in  the  female  zone  of  the 
crown.  The  branches  with  one  top  are  most  suitable  for 
examining  the  relationship  between  the  seminiference 
and  the  shoot  growth.  For  the  study  of  crop  dynamics, 
on  the  other  hand,  the  branches  with  a  few  tops  must  be 
used.  These  branches  often  occur  in  the  lower  part  of  the 
female  zone. 

Sample  branches  of  the  tree  must  be  taken  from  three 
age  groups  (30,  60,  and  100  years  old).  The  extreme  val- 
ues of  the  crop  capacity  (the  number  of  cones  on  one 
shoot)  are  omitted  in  the  case  of  young  branches  for  the 
first  few  years  before  cone  bearing  and  in  the  case  of  old 
branches  for  the  last  tens  of  years  after  cone  bearing.  The 
extremes  are  ignored  because  of  the  absence  of  cone 


203 


Figure  2 — Traces  of  generative  and  vegetative 
organs  on  cross-sectional  saw  cuts  of  branches. 
X  3:  (a)  cone  that  abscissed  before  pollination, 
(b)  cone  that  abscissed  after  pollination,  (c)  cone 
that  abscissed  during  maturation,  (d)  mature 
cone,  (e)  lateral  shoot,  (f)  brachyblast. 


traces.  The  information  is  taken  into  account  within  the 
number  series,  for  it  reflects  nonseed  years.  The  traces 
are  hardly  visible  at  this  time  because  of  the  predominant 
abscission  of  1-year-old  conelets  caused  by  early  frosts  or 
the  lack  of  pollination. 

No  less  than  three  to  five  200-  to  250-year-old  trees  of 
the  first  or  second  Kraft's  class  are  to  be  selected  from  one 
sample  area  for  the  study  of  sample  branches. 

A  retrospective  study  of  the  seminiference  of  Siberian 
cedar  pine  on  the  basis  of  the  proposed  method  makes  it 
possible  to  assess  the  effect  of  the  reproductive  activity  of 
trees  on  the  annual  ring  structure  and  size,  to  follow  the 
crop  cyclicity,  to  elaborate  the  ways  of  the  long-term  and 
super-long-term  forecasting  of  crops,  and  to  construct  an 
ontogenetic  model  of  the  relationship  between  the  growth 
and  seminiference  of  shoots.  This  method  appears  appli- 
cable to  other  coniferous  plants  and,  first  and  foremost, 
to  Pinaceae. 

REFERENCES 

Gorchakovsky,  P.  L.  1958.  New  issues  in  the  method  of  in- 
vestigation of  the  seminiference  dynamics  of  coniferous 
trees.  Bot.  Zhourn.  43(10):  1445-1459.  [In  Russian]. 

Iroshnikov,  A.  I.  1963.  The  cone-bearing  of  Siberian  cedar 
pine  in  western  Sayani.  In:  Cone-bearing  of  Siberian 
cedar  pine  in  eastern  Siberia;  Moscow:  Publ.  of  USSR 
Academy  of  Science:  104-119.  [In  Russian]. 

Korchagin,  A.  A.  1960.  Methods  of  account  of  seminiference 
of  trees  and  forest  communities.  In:  Field  Geobotany; 


Moscow-Leningrad:  Publication  of  USSR  Academy 
of  Science:  Vol.  2:  41-133.  [In  Russian]. 

Nekrasova,  T.  P.  1957.  On  the  methods  of  investigation 
of  cone-bearing  dynamics  in  coniferous  trees.  Izv.  Vost. 
Fil.  Akad.  Nauk  SSSR.  6:  138-145.  [In  Russian]. 

Nekrasova,  T.  P.  1972.  Biological  principles  of  seminif- 
erence of  Siberian  cedar  pine.  Novosibirsk,  Nauka: 
272  p.  [In  Russian]. 

Nesterov,  N.  S.  1914.  On  the  problem  of  the  methods  of 
investigation  of  cone-bearing  of  trees.  Forest-Industrial 
Vestnik.  Vol.  26.  [In  Russian]. 

Renvall,  A.  1912.  Die  periodischen  Erscheinungen  der 
Reproduction  der  Kiefer  an  der  polaren  Waldgrenze. 
Acta  Forest  Fenn.  22(3):  130  p. 

Trofimova,  Z.  I.  1953.  Determination  of  pine  crops  by  bio- 
logical method.  For.  Econ.  1:  67-69.  [In  Russi£Ui]. 

Vorobjev,  V.  N.  1964.  The  cone-bearing  of  Siberian  cedar 
pine  in  altitude  subzones  of  north-eastern  Altai.  Izv. 
Sib.  Otd.  Akad.  Nauk  SSSR,  Ser.  Biol.-Med.  Nauk. 
12(3):  86-90.  [In  Russian]. 

Vorobjev,  V.  N.  1974.  Peculiarities  of  cone-bearing  of 
Siberian  cedar  pine  in  mountain  conditions.  In:  Biology 
of  seed  reproduction  of  coniferous  trees  of  western 
Siberia.  Novosibirsk,  Nauka:  15-70.  [In  Russian]. 

Vorobjev,  V.  N.  1979.  A  method  of  retrospective  study 
of  seminiference  dynamics  in  Pinus  sibirica  Du  Tour 
(Pinaceae).  Bot.  Zhourn.  64(7):  971-974.  [In  Russian]. 

Vorobjev,  V.  N.;  Vorobjeva,  N.  A.;  Goroshkevich,  S.  N. 
1989.  Growth  and  sex  of  Siberian  cedar  pine. 
Novosibirsk,  Nauka.  167  p.  [In  Russian]. 


204 


1^  \\ 

Importance  to  Wildlife 


International  Workshop 
St.  Moritz  1 992 

^  — ^ 


205 


NUTRITIVE  VALUE  OF  WHITEBARK 
PINE  SEEDS,  AND  THE  QUESTION  OF 
THEIR  VARIABLE  DORMANCY 

Ronald  M.  Lanner 
Barrie  K.  Gilbert 


Abstract— Seeds  of  whitebark  pine  (Pinus  albicaulis)  are  eaten 
year-round  by  their  disperser,  Clark's  nutcracker  (Nucifraga 
Columbiana),  and  may  be  the  only  food  that  is  fed  to  their  young. 
They  are  also  preferentially  larder-hoarded  by  the  red  squirrel 
{Tamiasciurus  hudsonicus);  and  are  eaten  in  large  quantities  by 
black  bears  (Ursus  americanus)  and  grizzly  bears  {U.  arctos),  es- 
pecially prior  to  hibernation. 

Analyses  show  whitebark  pine  seeds  have  a  mean  seed  weight 
of  119  mg,  of  which  kernels  comprise  52.5  percent.  Kernel  content 
was  approximately  21  percent  carbohydrate,  21  percent  protein, 
52  percent  fat,  3  percent  ash,  and  3  percent  water.  Major  mineral 
nutrients  present  were  Cu,  Zn,  Fe,  Mn,  Mg,  and  Ca.  Sixteen  amino 
acids  were  present,  with  glutamic  acid,  lysine,  and  arginine  espe- 
cially abundant.  Major  fatty  acids  were  linoleic,  oleic,  and  y-linolenic, 
with  lesser  amounts  of  11  others.  The  unsaturated  fraction  was 
92  percent. 

The  variable  period  of  dormancy  is  analyzed  in  terms  of  its  con- 
tribution to  fitness  in  a  variable  environment,  its  phylogenetic 
distribution  within  Pinus  subg.  Strobus  and  subsect.  Cembrae, 
and  its  effect  in  retaining  nutcracker  populations  by  broadening 
the  food  base.  The  food  base  is  broadened  by  the  interaction  of 
memory  duration,  seed  crop  frequency,  and  foraging  on  buried 
dormant  seeds  whose  location  is  betrayed  by  presence  of  germi- 
nants.  This  may  explain  the  excessive  caching  behavior  of  nut- 
crackers and  may  reduce  the  frequency  of  irruptions  during  times 
of  food  shortage.  Even  when  seed  crops  are  sporadic,  variable  dor- 
mancy can  make  seeds  available  to  nutcrackers  in  most  years,  and 
as  early  as  the  first  year  after  dispersal  birds  can  forage  on  seeds 
cached  by  other  birds. 


Whitebark  pine  (Pinus  albicaulis  Engelm.)  is  the  only 
North  American  member  of  Pinus  subsect.  Cembrae,  the 
stone  pines.  The  species  is  found  in  subalpine  habitats  in 
the  Northern  Rocky  Mountains  of  Canada  and  the  United 
States,  and  in  the  Cascade  Mountains-Sierra  Nevada  sys- 
tem (Critchfield  and  Little  1966).  Because  whitebark  pine 
usually  grows  in  inaccessible  areas  and  has  relatively  low 
commercial  value,  it  was  long  neglected  by  researchers. 
But  in  recent  years,  due  largely  to  interest  in  corvid-pine 
mutualisms  stimulated  by  a  seminal  review  (Turcek  and 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Ronald  M.  Lanner  is  Professor,  Department  of  Forest  Resources,  and 
Barrie  K.  Gilbert  is  Senior  Scientist,  Department  of  Fisheries  and  Wildlife, 
both  at  Utah  State  University,  Logan,  UT. 


Kelso  1968),  it  has  received  greatly  increased  attention. 
Much  of  this  attention  has  focused  on  the  important  role 
of  whitebark  pine  seed  in  the  biology  of  animals  that  in- 
habit the  whitebark  pine  forest. 

It  is  the  large  Upid-rich  "nut"  of  this  pine  that  attracts  its 
seed  disperser,  Clark's  nutcracker  (Nucifraga  columbiana) 
(Tomback  1978).  Evidence  indicates  that  the  pine  can  reU- 
ably  regenerate  only  from  seeds  cached  in  the  soil  by  its  nut- 
cracker mutualist  (Hutchins  and  Lanner  1982),  and  that 
these  seeds  are  the  sole  nutritional  resource  for  nutcrackers 
throughout  their  juvenile  stage  (Vander  Wall  and  Hutchins 
1983).  Further,  it  has  been  shown  that  whitebark  pine 
seeds  are  a  preferred  food  of  the  red  squirrel  (Tamiasciurus 
hudsonicus),  which  larder-hoards  great  quantities  of  cones 
on  and  below  the  groimd  (Hutchins  and  Lanner  1982).  Both 
black  bears  (Ursus  americanus)  and  grizzly  bears  (U.  arctos) 
raid  the  squirrel  middens  in  order  to  gorge  themselves  with 
pine  nuts  both  before  and  after  hibernation  (Kendall  1983; 
Mattson  and  Jonkel  1990).  Thus  the  whitebark  pine  seed 
is  of  obviously  great  nutritional  importance  to  at  least  five 
organisms  of  its  ecosystem:  Clark's  nutcracker,  the  red 
squirrel,  black  and  grizzly  bears,  and  whitebark  pine  itself, 
for  whose  benefit  the  seed  originally  evolved.  It  is  therefore 
of  more  than  passing  interest  to  know  the  nutrient  content 
of  this  valued  seed. 

Pubhshed  analyses  of  whitebark  pine  nuts  have  stressed 
their  high  caloric  value,  which  is  mainly  a  function  of  high 
lipid  content.  Emphasis  on  caloric  content  appefirs  to  be  a 
response  to  the  current  ecological  fixation  on  energetics  of 
food  resources,  almost  to  the  exclusion  of  other  food  charac- 
teristics. Thus  Lanner  (1982)  has  documented  the  parti- 
tioning of  energy  in  nuts  and  other  cone  components,  and 
Hutchins  and  Lanner  (1982)  have  reported  changes  in  the 
caloric  value  of  whitebark  pine  seeds  as  the  seeds  mature, 
but  other  aspects  of  the  nutritional  qualities  of  these  seeds 
have  been  neglected. 

Another  characteristic  of  the  whitebark  pine  seed  that 
requires  attention  is  its  variable  dormancy.  Seeds  of  several 
Cembrae  species  have  been  reported  to  delay  germination 
until  the  second  or  third  year  following  dispersal  (USDA 
FS  1974).  This  has  been  attributed  to  the  immaturity  of 
embryos  at  the  normal  germination  date,  and  has  been 
blamed  for  low  germination  rates  of  whitebark  pine  seeds 
in  laboratory  tests  (Pitel  and  Wang  1990).  Kozlowski  and 
Gunn  (1972)  have  suggested  that  germination  delays  ensure 
estabhshment  "even  though  the  early  germinants  failed  to 
survive  severe  environmental  stresses  such  as  droughts  or 
severe  frosts,"  but  they  offer  no  data  to  support  this  intu- 
itively attractive  idea.  According  to  McCaughey  (1992), 
whitebark  pine  seeds  of  a  cohort  displayed  germination 


206 


dming  the  first  (11  percent),  second  (45  percent),  third 
(11  percent),  and  fourth  (percent  not  specified)  postdis- 
persal  years. 

The  objectives  of  this  paper  are  to  document  the  nutri- 
tional characteristics  of  whitebark  pine  seeds,  and  to  sug- 
gest some  possible  ramifications  of  the  variable  dormancy 
habit. 

SEED  COMPOSITION  ANALYSES 

Analyses  were  made  of  whitebark  pine  seeds  of  the  1989 
crop  fi-om  around  Gardiner,  MT  (2,635  m  elevation),  provided 
by  Dr.  Ward  McCaughey  of  the  Intermountain  Research  Sta- 
tion (USDA  Forest  Service),  Bozeman,  MT. 

Protein  determination  was  made  by  the  Kjeldahl  method, 
fat  determination  by  the  chloroform  methanol  method  of 
Hubbard  (1977),  and  mineral  assays  by  atomic  absorption 
by  Drs.  Arthiir  Mahoney  and  Dejia  Zhang,  Department  of 
Nutrition  and  Food  Sciences,  Utah  State  University.  Five 
samples  were  used  for  these  determinations.  Amino  acid 
analyses  of  nut  protein  were  made  by  Dr.  Rod  Brown,  De- 
partment of  Nutrition  and  Food  Sciences,  Utah  State  Uni- 
versity, using  a  Beckmann  6300  amino  acid  analyzer.  Four 
nms  were  made:  data  presented  here  are  the  means  of  those 
runs.  Fatty  acid  composition  of  kernel  fats  was  provided  by 
Laurence  G.  Cool  of  Dr.  E.  Zavarin's  laboratory  at  the  For- 
est Products  Laboratory,  University  of  Cahfomia,  Richmond, 
CA,  and  was  determined  by  gas  chromatography. 

Results 

Kernel  Composition — Mean  values  of  20-seed  samples 
yielded  the  following  seed  coat  versus  kernel  (female  game- 
toph5^e  plus  embryo)  breakdown:  seed  coat,  47.5  percent, 
0.0568  g;  kernel,  52.5  percent,  0.0626  g.  Kernels  consisted 
of  water  (3.5  percent),  ash  (2.8  percent),  carbohydrates 
(21  percent)  [calculated  value],  protein  (21  percent),  and 


Table  1 — Amino  acids  identified  in  kernel 
protein  of  whitebark  pine  in  order 
of  abundance  (mean  of  four  runs) 


Amino 

Nanomoles 

acid 

per  mg 

Glutamic  acid 

0.729 

Lysine 

.666 

Arginine 

.554 

Aspartic  acid 

.422 

Glycine 

.386 

Valine 

.380 

Alanine 

.373 

Leucine 

.369 

Serine 

.363 

Proline 

.352 

Isoleucine 

.245 

Phenylalanine 

.192 

Tyrosine 

.168 

Threonine 

.147 

Histidine 

.126 

Methionine 

.038 

Table  2 — Comparative  abundance  of  amino  acids  identified  in 
kernel  protein  of  three  pine  species^ 


Amino 
acid' 

P.  albicaulis 

P.  edulis 

P.  pinea 

^1 1 1 

1  An 

0  / 

ft7 

1  vo 
LYb 

y  1 

1  Q 

A 

Ano 

fO 

1  nn 

1  UU 

1  nn 

ACD 

RQ 
OO 

A7 

vaLY 

til 

OR 

VAL 

07 

Al  A 
MLA 

Dl 

01 

97 

1  CI  1 

01 

O  / 

Ol 

CCD 
OtPl 

OU 

00 

OO 
r  r 

rnU 

AO 

OQ. 

loU 

'>A 

o4 

OCi 

on 

Due 

rn  t 

on 

on 

1  Yn 

23 

H  n 

i» 

THR 

20 

16 

16 

HIS 

17 

12 

12 

MET 

5 

9 

9 

CYS 

9 

9 

TRY 

6 

1 

'Species  and  references:  P.  albicaulis  (this  paper),  P.  edulis  (McCarthy  and 
Matthews  1984),  P.  pinea  (McCarthy  and  Matthews  1984). 

^Amino  acids  not  previously  mentioned;  CYS  =  cystine,  TRY  =  tryptophan. 


fat  (52  percent).  Atomic  absorption  analysis  disclosed  the 
following  mineral  composition,  in  micrograms  per  gram  of 
ash:  Cu,  17.5;  Zn,  66.9;  Fe,  35.7;  Mn,  132;  Mg,  17.5;  Ca,  51.9. 

Amino  Acids — Sixteen  amino  acids  were  identified  in  the 
whitebark  pine  kernel  samples.  The  most  abimdant  were 
glutamic  acid,  lysine,  arginine,  and  aspartic  acid  (table  1). 
These  ranged  from  0.42  to  0.73  nmol/mg.  Also  abxmdant, 
ranging  from  0.35  to  0.39  nmol/mg  were  glycine,  valine,  ala- 
nine, leucine,  serine,  and  proline.  Lesser  quantities  were 
found  of  isoleucine,  phenylalanine,  tyrosine,  threonine,  and 
histidine;  and  least  of  methionine  (table  1). 

Amino  acid  quantities  have  been  expressed  in  the  Htera- 
ture  in  various  ways.  Therefore,  to  compare  the  amino  acids 
of  whitebark  pine  with  those  of  other  species,  we  have  given 
the  most  abimdant  amino  acid  of  each  species  the  index 
value  of  100,  and  aU  others  are  scaled  appropriately  within 
each  species.  Thus,  while  the  most  abimdant  amino  acid 
(100)  of  whitebark  pine  is  glutamic  acid,  that  oiPinus  edulis 
and  P.  pinea  is  arginine  (table  2). 

Fatty  Acids — Fifteen  fatty  acids  were  identified  on  the 
chromatograph.  The  most  abimdant  of  these  were  the  poly- 
unsaturated linoleic  acid  and  monounsaturated  oleic  acid, 
which  comprised  46.9  percent  and  30.6  percent  of  the  total 
fatty  acid  methyl  esters,  respectively.  The  other  components 
are  present  in  much  smaller  proportions  (table  3).  The  total 
fraction  of  imsaturated  fatty  acids  was  92.0  percent,  32.1 
percent  monoimsaturated,  and  59.9  percent  polyunsaturated. 

Discussion 

Due  to  differences  in  analytical  techniques,  and  variation 
in  the  multipHer  used  in  determining  protein  content,  the 
pubUshed  fractions  of  fats,  proteins,  carbohydrates,  and 
ash  in  nut  kernels  are  quite  variable. 


207 


Table  3 — Fatty  acids  identified  in  kernel  fat  of 
whitebark  pine  in  order  of  abundance 


Total  fatty  acid 

Fatty  acid 

methyl  esters 

Percent 

Linoleic 

46.9 

Oleic 

30.6 

y-Linolenic 

9.9 

Palmitic 

4.2 

Stearic 

2.7 

1 1  -Eicosenoic 

1.4 

y-Linoleic 

1.2 

5-,  11-,  14-Eicosatrienoic 

.8 

1 1  -,  1 4-Eicosadienoic 

.7 

Arachidic 

.6 

Linolenic 

.3 

DOl  lc?l  IIU 

o 

Vaccenic 

.1 

5-,  1 1  -Eicosadienoic 

.1 

Unknowns 

.3 

Total 

100.0 

Whitebark  pine's  protein  content  of  21  percent  compares 
with  pinyon  pine  values  in  the  range  of  10  percent  (Pinus 
monophylla)  to  19  percent  (P.  cembroides)  (Lanner  1981); 
25  percent  in  P.  sabiniana,  and  31  percent  in  P.  pinea 
(Farris  1983). 

The  proteins  in  pine  nuts  are  almost  entirely  seed  storage 
proteins  of  the  megagametophyte.  Following  the  hydrolysis 
of  these  proteins,  the  component  amino  acids  are  exported 
into  the  growing  points  of  the  embryo  (Lammer  and  GifiFord 
1989).  GifiFord  (1988)  has  shown  there  is  considerable  varia- 
tion in  electrophoretic  characteristics  of  seed  proteins  of 
several  species  (including  P.  albicaulis);  and  Schirone  and 
others  (1991)  have  shown  such  variation  to  have  taxonomic 
significance.  It  is  therefore  surprising  to  see  the  almost- 
identical  gmaino  acid  profiles  of  P.  pinea  and  P.  edulis  seed 
proteins  (table  2).  These  species  are  in  separate  subgenera 
of  Pinus  (Critchfield  and  Little  1966)  and  would  be  expected 
to  have  proteins  of  quite  different  composition. 

Foiu"  common  amino  acids  were  missing  fi-om  whitebark 
pine  seed  protein.  Hydroxyproline's  absence  can  probably 
be  attributed  to  a  lack  of  that  amino  acid  in  embryo  and 
storage  protein,  and  its  restriction  to  seed  coats  and  peri- 
carp material  (Van  Etten  and  others  1967).  Tr3T)tophan  is 
generally  a  minor  component  of  plant  proteins  (White  and 
others  1959)  and  can  be  destroyed  in  analysis  (Farris  1983). 
Neither  cystine  nor  cysteine  were  identified  in  whitebark 
pine,  though  the  former  is  found  in  small  quantities  in 
P.  edulis  and  P.  pinea  (table  2).  The  lack  of  cystine  and  the 
low  quantity  of  methionine  indicate  that  sulfur-containing 
amino  acids  are  of  little  importance  to  development  of  white- 
bark pine  seedlings. 

Perhaps  the  most  striking  difference  between  seed  protein 
of  whitebark  pine  and  the  other  species  in  table  2  is  its  great 
abundance  of  lysine.  Lysine,  an  essential  amino  acid  for 
human  growth,  is  notoriously  low  in  cultivated  members  of 
the  Graminae.  Feeding  experiments  conducted  with  dark- 
eyed  juncos  iJunco  hyemalis)  by  Parrish  and  Martin  (1977) 


have  established  the  essentiality  of  lysine  in  maintaining 
a  positive  nitrogen  balance  in  that  bird.  Additional  amino 
acids  found  essential  for  poultry  are  arginine,  glycine,  £ind 
histidine,  all  of  which  are  found  in  whitebark  pine  nuts 
(table  1).  In  the  absence  of  data  specific  to  the  genus  Nuci- 
fraga,  we  speculate  that  these  nuts  contain  all  the  amino 
acids  essential  to  nutcrackers. 

According  to  Van  Etten  and  others  (1967)  the  nutrition- 
ally essential  amino  acid  requirements  for  swine  are  similar 
to  those  of  humans.  If  the  requirements  for  bears  are  also 
reasonably  similar,  then  whitebark  pine  protein,  as  well  as 
its  fat,  may  be  of  great  dietary  significance  for  bears.  Bears 
are  more  likely  to  satisfy  their  protein  needs  firom  a  wider 
variety  of  foodstuffs  than  nutcrackers,  however. 

The  fat  content  of  whitebark  pine  nuts  in  this  study  (52 
percent)  is  qmte  similar  to  that  of  other  stone  pines  as  deter- 
mined by  Tikhomirov  (1939):  P.  cembra,  59.9  percent  and 
P.  pumila,  59.4  percent.  Fat  values  for  P.  sibirica  vary  over 
the  range  51  to  75  percent  (Shimanjoik  1963).  The  fat  con- 
tent of  pinyon  pines  (Lsmner  1981)  ranges  fi'om  23  percent 
in  P.  monophylla  to  62  to  71  percent  in  P.  edulis.  Farris 
(1983)  reports  that  P.  sabiniana  kernels  contain  49  percent 
of  fat,  and  P.  pinea  contains  47  percent. 

Tikhomirov  (1939)  reported  the  fat  content  of  P.  koraien- 
sis  to  be  52.5  percent,  but  recent  work  by  Yoon  and  others 
(1989)  reports  a  crude  lipid  fi'action  of  72.5  percent,  which 
reduced  to  70.4  percent  when  purified. 

The  most  notable  features  of  the  whitebark  pine  seed's 
fatty  acids  are  the  preponderance  of  oleic  (30.6  percent)  and 
linoleic  (46.9  percent)  acids.  These  are  both  unsaturated 
fatty  acids,  and  it  appears  they  are  commonly  the  major 
ingredients  of  the  seed  fatty  acids  (f.a.)  of  Pinus.  Thus  they 
comprise  28.4  percent  and  47.9  percent  of  the  f.a.  of  P.  kor- 
aiensis,  40.0  percent  and  41.9  percent  of  the  f.a.  of  P.  edulis 
(R.  M.  Lanner,  impubl.  data),  and  42.5  percent  and  42.0  per- 
cent of  the  f.a.  of  P.  monophylla,  respectively  (R.  M.  Lanner, 
impubl.  data).  According  to  Yoon  and  others  (1989)  the  f.a. 
of  P.  koraiensis  are  8.75  percent  saturated,  29.41  percent 
monounsaturated,  and  61.64  percent  poljomsaturated.  Pinus 
edulis  and  P.  monophylla  have  over  80  percent  polyunsatu- 
rates. The  nut  of  P.  sabiniana  was  found  by  Farris  (1983) 
to  be  50.5  percent  oleic  and  45.2  percent  linoleic  acids — 
over  95  percent  polyimsaturated. 

In  addition  to  containing  very  high  ratios  of  imsaturated 
to  saturated  fatty  acids,  pine  nuts  are  also  fi'ee  of  cholesterol 
(McCarthy  and  Matthews  1984).  Yoon  and  others  (1989) 
point  out  that  P.  koraiensis  nuts  have  long  been  used  in 
Korean  medicine,  and  the  nut  oil  of  P.  cembra  had  medici- 
nal uses  in  Poland  during  the  17th  century  (Bialobok  1975). 
Yoon  and  others  point  out  that  P.  koraiensis  seeds  contain 
omega-5  fatty  acids,  the  functions  of  which  in  the  himian 
body  are  unknown.  They  suggest  that  these  f.a.  may  be 
beneficial  to  health,  and  should  be  studied  in  relation  to 
lipid  and  glucose  metabolism. 

VARIABLE  DORMANCY 

The  variable  dormancy  that  characterizes  whitebark 
pine  seeds  has  been  viewed  as  a  bothersome  habit  that 
makes  the  nurseryman's  task  more  difiScult.  However,  it 
could  well  be  a  trait  that  has  an  important  biological  func- 
tion and  that  has  been  naturally  selected  to  perform  that 


208 


Table  4^Number  of  years  during  a  15-year  period  in  which  seeds  will  be  germinating,  under  assumptions  of  different  seed-crop  intervals,  and 
of  dormancy  varying  from  0  to  3  years^ 


Period  of  seed 

dormancy  Interval  between  successive  seed  crops 


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

0 

15 

8 

5 

4 

3 

3 

3 

2 

2 

2 

2 

2 

2 

2 

1 

1 

15 

15 

10 

8 

6 

6 

5 

4 

4 

4 

4 

4 

4 

3 

2 

2 

15 

15 

15 

12 

9 

9 

7 

6 

6 

6 

6 

6 

5 

4 

3 

3 

15 

15 

15 

15 

12 

11 

9 

8 

8 

8 

8 

7 

6 

5 

4 

'Seeds  dormant  for  0  years  all  germinate  the  spring  following  dispersal.  Seeds  domiant  more  than  0  years  germinate  in  part  in  the  spring  following  disperal;  and 
in  part  during  each  additional  spring  for  1 ,  2,  or  3  years. 


function.  Here  we  postulate  and  examine  several  possible 
evolutionary  functions  for  variable  dormancy. 

Increased  Fitness 

According  to  this  hypothesis,  the  conditions  met  by  ger- 
minating seeds  may  vary  greatly  from  year  to  year.  There- 
fore a  cohort  of  seeds  dispersed  in  fall  of  the  year  n  may  be 
confronted  by  disadvantageous — or  even  catastrophic — 
conditions  in  spring  of  n+1.  Such  conditions  could  result 
from  severe  frost  or  drought  (Kozlowski  and  Gunn  1972). 
If  so,  it  might  be  to  the  plant's  advantage  to  delay  germi- 
nation of  some  seeds  until  spring  of  n+2  or  even  n+3  or  be- 
yond. If  the  temporal  spacing  of  seed  crops  and  the  lengths 
of  the  variable  dormancy  periods  are  matched  properly,  par- 
tial seed  cohorts  could  be  germinating  nearly  every  year. 
Under  such  a  sceneirio  there  would  be  a  high  probability  of 
some  seeds  germinating  during  the  years  of  most  favorable 
conditions. 

As  an  example,  assimie  a  pine  stand  bearing  seeds  in  the 
first  year  of  an  arbitrarily  chosen  15-year  period.  If  seed 
crops  reoccur  every  4  years  (in  years  5,  9,  and  13  of  this 
15-year  period),  and  if  germination  occiu-s  only  in  the  spring 
foUovdng  dispersal  (spring  of  years  2,  6, 10,  and  14),  then 
there  will  be  4  years  during  the  15-year  period  when  seed- 
lings can  become  established  (table  4).  However,  if  some 
of  those  seeds  remain  dormant  for  variable  lengths  of  time, 
germinating  not  only  in  the  spring  following  dispersal,  but 
in  the  next  three  springs  as  well,  seedlings  can  become  es- 
tablished during  all  15  years  of  the  period  (table  4).  Even 
if  seed  crops  were  spaced  at  10-year  intervals,  this  level  of 
variable  dormancy  would  produce  seedlings  in  8  of  the  15 
years.  Thus  it  is  apparent  that  variable  dormancy  can  neu- 
tralize the  effects  of  infrequent  crop  years  by  creating  a 
steady  stream  of  germinating  seeds  during  and  between 
crop  years. 

If  variable  dormancy  increases  whitebark  pine's  fitness  by 
increasing  the  frequency  of  germination  years,  one  might  ex- 
pect associated  conifers — especially  other  pines — ^to  show  the 
same  dormancy  behavior.  Apparently,  however,  this  seems 
not  to  be  the  case.  If  we  assume  that  stratification  treat- 
ments prescribed  to  overcome  seed  dormancy  are  an  index 
of  a  seed's  behavior  in  nature,  it  is  apparent  that  the  subal- 
pine  conifers  of  western  North  America  show  no  consistency 
(table  5).  This  does  not  mean  that  whitebark  pine  derives 
no  benefits  from  being  able  to  provide  germinants  in  most 


years,  but  perhaps  that  only  whitebark  pine  (and  maybe 
P.  monticola)  enjoys  those  benefits.  Apparently  variable 
dormancy  is  not  a  requirement  for  success  of  whitebark 
pine's  associates  in  the  subsdpine  zone.  In  fact,  variable 
dormancy  might  be  detrimental  to  these  other  conifers, 
because,  except  for  P.  flexilis,  they  have  wind-dispersed 
seeds  that  lie  on  the  soil  surface,  where  they  are  vulner- 
able to  predation.  Buried  seeds  of  whitebark  pine  (and  lim- 
ber pine)  would  be  more  likely  to  survive  an  additional  year 
or  more  of  "waiting"  to  germinate. 

If  variable  dormancy  is  a  genetically  controlled  charac- 
ter, it  would  be  important  to  learn  its  mode  of  inheritance. 
For  example,  do  all  the  seeds  of  a  tree  exhibit  the  same 
dormancy  behavior,  as  would  be  expected  in  a  maternally 
transmitted  characteristic  of  seed-coat  or  megagametophytic 
tissue;  or  do  they  segregate,  as  would  be  expected  if  inher- 
itance from  both  parents  is  manifested  in  variable  embryo 
behavior?  If  the  latter,  is  genetic  control  due  to  major  gene 
effects,  or  is  control  achieved  multigenically?  Seed  tests 
would  need  to  be  made  with  the  products  of  controlled 
crosses  to  unravel  this  and  similar  questions,  and  careful 
anatomical  studies  would  have  to  be  made  to  better  imder- 
stand  embryo  growth  and  development. 


Table  5 — Recommended^  cold  stratification  periods  for  seeds  of 
subalpine  conifers  of  western  North  America  and  their 
overlap  with  that  of  Pinus  albicaulis 

Overlap  of 

Stratification  P.  albicaulis 

Species  period  stratification  period 

Days  Percent 


Pinus  albicaulis 

90-120 

P.  monticola 

30-120 

100 

P.  flexilis 

21  -  90 

3 

P.  aristata 

0 

0 

P.  balfouriana 

90 

3 

P.  contorta  var.  latifolia 

30-  56 

0 

P.  contorta  var.  murrayana 

20-  30 

0 

Tsuga  mertensiana 

90 

3 

Picea  engelmannii 

0 

0 

Abies  lasiocarpa 

28 

0 

A.  amabilis 

21  -  28 

0 

A.  magnifica 

28-  42 

0 

In  USDA  FS1974. 


209 


Phylogenetic  Legacy 

Perhaps  whitebark  pine  exhibits  variable  dormancy  sim- 
ply because  the  trait  is  conmion  to  a  group  of  related  species 
that  includes  whitebark  pine.  To  determine  this,  we  have 
displayed  in  table  6  the  available  stratification  requirements 
of  pines  of  Pinus  subg.  Strobus  by  subsection. 

Species  characterized  (in  USDA  FS  1974)  as  "suspected 
of  having  immature  embryos  at  the  time  of  collection"  or  of 
exhibiting  "extreme  dormancy"  are  identified  by  superscript 
nvunbers.  The  densest  concentration  of  superscripts  is  in 
the  stone  pines — subsection  Cembrae — in  which  all  species 
require  more  than  60  days  of  stratification  to  germinate 
(table  6).  Thus  it  seems  likely  that  variable  dormancy  is 
not  just  an  albicaulis  trait,  nor  a  broad  Strobus  trait,  but 
a  subsection  Cembrae  trait.  This  is  consistent  with  my  ear- 
lier suggestion  (Lanner  1980)  that  the  Cembrae  pines  are 
of  monophyletic  origin. 

Nutcracker  Retention 

The  importance  of  variable  dormancy  may  lie  in  its  ef- 
fects on  whitebark  pine's  seed  disperser,  Clark's  nutcracker. 
Whitebark  pine  depends  on  the  nutcracker  for  seed  dispersal 
and  seedling  establishment  (Hutchins  and  Lanner  1982), 
so  it  is  in  the  long-term  interest  of  the  pine  to  retain  nut- 
cracker populations  in  its  ecosystem.  We  suggest  that  vari- 
able dormancy  can  have  that  effect. 

The  following  scenario  illustrates  our  argument.  A  mast 
year  occurs  in  whitebark  pine  (year  n).  The  local  nutcracker 
population  frantically  harvests  and  caches  several  times 
as  many  seeds  as  are  required  to  ensure  its  own  survival 
and  that  of  its  offspring  imtil  the  next  fall  (Hutchins  and 
Lanner  1982;  Tomback  1978;  Vander  Wall  and  Balda  1977). 
Throu^out  the  fall,  the  nutcrackers  recache  their  buried 
seeds  (Hutchins  and  Lanner  1982).  The  next  summer,  in 
year  n+1,  some  seeds  germinate.  The  birds  now  have  three 
sources  of  food — their  own  caches,  fi'om  which  they  continue 
to  feed;  the  seeds  now  germinating  and  therefore  visible; 
and  the  still-dormant  seeds  fi'om  the  same  caches,  whose 
locations  are  betrayed  by  the  n+1  germinants  (Vander  Wall 
and  Hutchins  1983).  Each  bird  now  knows  not  only  the 
locations  of  all  its  own  caches,  but  also  the  caches  made 
by  other  birds  in  which  n+1  germinants  were  found.  Even 
if  there  is  no  crop  in  the  second  year,  the  bird's  9-month 
memory  (Balda  1992)  of  cache  location  will  allow  it  to  sur- 
vive until  the  third  year,  when  a  new  crop  may  be  present. 
Again,  sprouting  seeds  of  n+2  will  betray  the  locations  of 
remaining  caches.  If  there  is  again  no  crop,  or  only  a  light 
one,  the  nutcrackers  may  be  able  to  survive  by  recaching 
the  dormant  seeds  they  have  found  associated  with  germi- 
nating seeds.  This  assumes  that  another  9-month  period 
of  cache  remembrance  follows  each  caching  event.  Due  to 
the  overlapping  of  periods  of  availability  of  current-crop 
seeds  and  seeds  of  the  previous  crop  still  lying  dormant  up 
to  3  years  in  caches,  plus  those  germinating,  nutcrackers 
should  be  able  to  remain  in  situ  except  in  years  following 
several  successive  crop  failures. 

Such  a  scenario  could  not  occur  with  seeds  that  all 
sprouted  in  n+1.  With  variable  dormancy  it  appears  to  be 
a  strong  possibility;  however,  intensive  observation  and 
additional  data  will  be  required  to  determine  whether  it 


Table  6— Recommended^  cold  stratification  periods  for  seed  of 

some  pines  in  subgenus  Strobus,  and  their  overlap  with 
that  of  Pinus  albicaulis 


Species 


Stratification 
period 


Overlap  of 
P.  albicaulis 
stratification  period 


Days  Percent 

(Pinus) 

albicaulis^                       90-120  — 

subsection  Strobi 

flexilis^                           21-90  3 

monticola^                        30-120  100 

strobus                             60  0 

strobiformis^                     60-120  100 

lambertiana^                     60-  90  3 

peuce                             0-60  0 

parviflora^-^                         90  3 

griff ithii                            0-15  0 

subsection  Balfourianae 

aristata                             0  0 

balfouriana^                       90  3 

subsection  Cembroides 

cembroides                         0  0 

edulis                               0  0 

monophylla^                      28-  90  3 

quadrifolia                           0  0 

subsection  Cembrae 

cembra^'^                        90-270  100 

s/b/r/ca"                          60-  90  3 

pumila^                         120-150  3 

koraiensis^'^                        90  3 

'In  USDA  FS  1974. 

^Characterized  as  exhibiting  "extreme  dormancy"  (requiring  more  than  60 
days  of  stratification  in  USDA  FS  1974). 

'Characterized  as  "suspected  of  having  immature  embryos  at  the  time  of 
collection"  in  USDA  FS  1974. 


is  realistic.  Our  intention  in  speculating  is  to  stimulate 
fiirther  study  of  this  possibility.  All  of  this  illustrates  that 
to  understand  whitebark  pine's  variable  seed  dormancy, 
the  nutcracker  too  must  be  considered.  As  pointed  out  by 
Lanner  (1980):  "The  ecology  and  evolution  of  the  wingless- 
seeded  'Bird  Pines'  can  only  be  understood  in  a  frame  of 
reference  that  includes  the  behavior  of  their  feathered  dis- 
persers."  The  bare  facts  of  the  mutualism  of  whitebark 
pine  and  Clark's  nutcracker  are  by  now  well  established. 
The  time  is  ripe  to  study  the  complexities  of  the  relation- 
ship, and  its  second-order  effects. 

REFERENCES 

Balda,  R.  P.  1992.  [Conversation  with  Ronald  M.  Lanner]. 
August  14.  Flagstaff,  AZ:  Northern  Arizona  University. 

Bialobok,  S.,  ed.  1975.  Stone-pine  (Pinus  cembra  L.).  Trans- 
lated from  Polish  for  U.S.  Department  of  Agriculture  and 
National  Science  Foundation.  National  Technicsd  Infor- 
mation Service  TT-73-54045.  173  p. 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  Misc.  Publ.  991. 


210 


Washington,  DC:  U.S.  Department  of  Agriciilture,  Forest 
Service.  97  p. 

Farris,  G.  J.  1983.  California  pignolia:  seeds  of  Pinus  sabini- 
ana.  Economic  Botany.  37:  201-206. 

Gifford,  D.  J.  1988.  An  electrophoretic  analysis  of  the  seed 
proteins  from  Pinus  monticola  and  eight  other  species  of 
pine.  Canadian  Journal  of  Botany.  66:  1808-1812. 

Hubbard,  W.  D.  1977.  Comparison  of  various  methods  for 
the  extraction  of  total  lipids,  fatty  acids,  cholesterol,  and 
other  sterols  from  food  products.  Journal  of  the  American 
Oil  Chemists'  Society.  54:  81-83. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  estabhshment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Kendall,  K  1983.  Use  of  pine  nuts  by  black  and  grizzly  bears 
in  the  Yellowstone  area.  In:  Meslaw,  Charles  E.,  ed.  Bears — 
their  biology  and  management.  West  Glacier,  MT:  Inter- 
national Association  for  Bear  Research  and  Management: 
166-173. 

Kozlowski,  T.  T.;  Gunn,  C.  R.  1972.  Importance  and  charac- 
teristics of  seeds.  In:  Kozlowski,  T.  T.,  ed.  Seed  biology. 
New  York:  Academic  Press:  1-20. 

Lammer,  D.  L.;  Gifford,  D.  J.  1989.  Lodgepole  pine  seed 
germination.  II.  The  seed  proteins  and  their  mobihzation 
in  the  megagametophyte  and  embryonic  axis.  Csmadian 
Journal  of  Botany.  67:  2544-2551. 

Lanner,  R.  M.  1980.  Avisin  seed  dispersal  as  a  factor  in  the 
ecology  and  evolution  of  limber  and  whitebark  pines. 
In:  Dancik,  B.  P.;  Higginbotham,  K  O.,  eds.  Sixth  North 
American  forest  biology  workshop  proceedings;  1980 
August  11-13;  Edmonton,  AB.  Edmonton,  AB:  University 
of  Alberta:  15-45. 

Lanner,  R.  M.  1981.  The  pinon  pine:  a  natural  and  cultural 
history.  Reno,  NV:  University  of  Nevada  Press.  208  p. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for  seed 
dispersal  by  Clark's  nutcracker.  Canadian  Journal  of 
Forest  Research.  12:  391-402. 

Mattson,  D.  J.;  Jonkel,  C.  1990.  Stone  pines  and  bears.  In: 
Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  Proceedings — 
sjnnposium  on  whitebark  pine  ecosystems:  ecology  of  a 
high-mountain  resoiu-ce;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermoimtain  Re- 
search Station:  223-236. 

McCarthy,  M.  A.;  Matthews,  R.  H.  1984.  Composition 
of  foods:  nut  and  seed  products.  Agric.  Handb.  8-12. 
Washington,  DC:  U.S.  Department  of  Agriculture,  Hu- 
man Nutrition  Information  Service.  137  p. 

McCaughey,  W.  1992.  [Letter  to  Ronald  M.  Lanner]. 
February  25.  1  leaf.  Bozeman,  MT:  U.S.  Department 


of  Agriculture,  Forest  Service,  Intermountain  Research 
Station. 

Parrish,  J.  W.,  Jr.;  Martin,  E.  W.  1977.  The  effect  of  dietary 
lysine  level  on  the  energy  and  nitrogen  balance  of  the 
dark-eyed  junco.  Condor.  79:  24-30. 

Pitel,  J.  A.;  Wang,  B.  S.  P.  1990.  Physical  and  chemical 
treatments  to  improve  germination  of  whitebark  pine 
seeds.  In:  Schmidt,  W.  C;  McDonald,  K  J.,  comps. 
Proceedings — sjmiposiiun  on  whitebark  pine  ecosystems: 
ecology  of  a  high-moxmtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  130-133. 

Schirone,  B.;  Piovesan,  B.;  Bellarosa,  R.;  Pelosi,  C.  1991.  A 
teixonomic  analysis  of  seed  proteins  in  Pinus  spp.  (Pina- 
ceae).  Plant  Systematics  and  Evolution.  178:  43-53. 

Shimanyuk,  A.  P.,  ed.  1963.  Fruiting  of  the  Siberian  stone 
pine  in  east  Siberia.  Translated  from  Russian  for  U.S. 
Department  of  Agricultural  and  National  Science  Foim- 
dation  by  Israel  Program  for  Scientific  Translations. 
Jerusalem.  204  p. 

Tikhomirov,  B.  A.  1939.  Khozyaistvenoye  ispol'zovaniye 
kedrovogo  stlannika.  Sov.  Sever.  4:  62-65. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. Living  Bird.  16:  123-161. 

Turcek,  F.  J.;  Kelso,  L.  1968.  Ecological  aspects  of  food 
transportation  and  storage  in  the  Corvidae.  Communica- 
tions in  Behavioral  Biology  Part  A.  1:  277-297. 

U.S.  Department  of  Agriculture,  Forest  Service.  1974.  Seeds 
of  woody  plants  in  the  United  States.  Agric.  Handb.  450. 
Washington,  DC:  U.S.  Department  of  Agricultiire,  Forest 
Service.  883  p. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of  the 
Clark's  nutcracker  and  the  pinon  pine  for  efficient  seed 
harvest  and  dispersal.  Ecological  Monographs.  47:  89-111. 

Vander  Wall,  S.  B.;  Hutchins,  H.  E.  1983.  Dependence  of 
Clark's  nutcracker,  Nucifraga  columbiana,  on  conifer 
seeds  dming  the  postfledging  period.  Canadian  Field 
Naturahst.  97:  208-214. 

Van  Etten,  C.  H.;  Kwolek,  W.  F.;  Peters,  J.  E.;  Barclay,  A.  S. 
1967.  Plant  seeds  as  protein  sources  for  food  or  feed.  Eval- 
uation based  on  amino  acid  composition  of  379  species. 
Journal  of  Agricultural  and  Food  Chemistry.  15: 1077-1089. 

White,  A.;  Handler,  P.;  Smith,  E.  L.;  Stetten,  D.  1959.  Prin- 
ciples of  biochemistry.  New  York:  McGraw-Hill  Book  Co. 
1149  p. 

Yoon,  T.-H.;  Im,  K-J.;  Koh,  E.  T.;  Ju,  J.-S.  1989.  Fatty  acid 
compositions  of  Pinus  koraiensis  seed.  Nutrition  Re- 
search. 9:  357-361. 


211 


BEAR  USE  OF  WHITEBARK  PINE 
SEEDS  IN  NORTH  AMERICA 


David  J.  Mattson 
Daniel  P.  Reinhart 


Abstract — Whitebark  pine  (Pinus  albicaulis)  seeds  are  an  im- 
portant high-quality  food  for  bear  populations  that  occupy  ecosys- 
tems with  continental  climates  south  of  the  United  States- 
Canada  border.  Availability  of  pine  seeds  affects  levels  of 
human-bear  conflict  and  bear  mortality.  In  most  areas  bears 
acquire  whitebark  pine  seeds  by  excavating  red  squirrel  (Tamia- 
sciurus  hudsonicus)  food  caches.  Squirrel  densities  appear  to 
limit  bear  use  of  pine  seeds  more  than  abundance  of  whitebark 
pine.  Thus  management  of  whitebark  pine  habitats  for  bear  use 
primarily  involves  management  to  favor  red  squirrels.  Bear  use 
of  pine  seeds  is  restricted  to  stands  >100  years  old,  and  can  per- 
sist for  an  additional  200  to  300  years.  Consequently  stand  rota- 
tions of  300  years  and  harvest  rates  of  3  percent  per  decade  are 
recommended.  Also,  because  of  potential  or  demonstrated  nega- 
tive impacts  associated  with  global  climate  change  and  white 
pine  blister  rust  {Cronartium  ribicola),  long-term  prospects  for 
whitebark  pine  should  be  part  of  considerations  when  managing 
habitats  of  bear  populations,  especially  insular  populations. 


Whitebark  pine  seeds  are  a  predictably  high-quality 
bear  food  given  their  relatively  leirge  size,  high  fat  con- 
tent, and  moderately  high  digestibility  (table  1).  Fat  con- 
tent is  perhaps  the  most  important  attribute  of  these 
seeds,  given  the  importance  of  body  fat  to  the  life  strategy 
of  bears  (Stirling  and  Derocher  1990)  and  the  greater  effi- 
ciencies of  body  fat  accimiulation  associated  with  high  di- 
etary fat  content  (Hadley  1985;  Schemmel  1976).  On  the 
other  hand,  use  of  whitebark  pine  seeds  especially  by  griz- 
zly bears  (Ursus  arctos  horribilis)  is  complicated  by  the  in- 
dehiscent  nature  of  whitebark  pine  cones  (Lanner  1990), 
and  the  consequent  reliance  of  most  bears  on  vdndthrow 
or  rodent  intermediaries  for  access  to  the  seeds  (Kendell 
1983;  Mattson  and  Jonkel  1990;  Taylor  1964). 

Black  (Ursus  americanus)  and  grizzly  bears  have  prob- 
ably made  use  of  pine  seeds  for  as  long  as  these  two  spe- 
cies have  coexisted  with  whitebark  pine,  approximately 
2  million  years  for  black  bears  and  12,000  years  for  griz- 
zly bears  (Axelrod  1986;  Km-ten  and  Anderson  1980).  Al- 
though aboriginal  Americans  were  probably  aware  of  this 
relationship,  the  first  written  records  of  bears  using 
whitebark  pine  seeds  did  not  appear  until  well  into  the 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

David  J.  Mattson  is  Wildlife  Biologist,  Interagency  Grizzly  Bear  Study 
Team,  Forestry  Sciences  Laborestry,  Montana  State  University,  Bozeman, 
MT  59717;  Daniel  P.  Reinhart  is  Management  Biologist,  National  Park 
Service,  Resource  Management,  Yellowstone  National  Park,  WY  82190. 


20th  century,  in  articles  by  authors  such  as  Cahalane 
(1947),  Taylor  (1964),  and  Tisch  (1961).  Details  of  the  re- 
lationship between  bears  suid  whitebark  pine  seeds  have 
only  recently  been  published,  focusing  on  interior  areas 
characterized  by  more  continental  climates  (Craighead 
and  others  1982;  Kendall  1983;  Mattson  and  Jonkel  1990; 
Mattson  and  others  1992, 1993).  In  this  paper  we  update 
information  presented  by  Mattson  and  Jonkel  (1990),  and 
further  elaborate  on  some  aspects  of  the  relationship  be- 
tween bears  and  whitebark  pine  in  North  America. 

PATTERNS  OF  BEAR  USE 
Geographic  Distribution 

Bears  consistently  make  use  of  whitebark  pine  seeds 
only  south  of  the  United  States-Ceinada  border,  where 
whitebark  pine  is  a  potentially  abundant  component  of 
subalpine  forests  (fig.  1).  However,  even  in  this  eirea  ma- 
jor differences  in  intensity  of  bear  use  are  evident,  associ- 
ated with  maritime  and  continental  climates. 

We  examined  published  bear  food  habits  studies  fi-om 
western  North  America  that  included  the  results  of  13,130 
bear  feces  examinations.  Our  analysis  clearly  showed 
that  the  consistency  and  relative  frequency  of  whitebark 
pine  seed  use  was  much  greater  (18.7  percent)  in  areas 


Table  1— Size,  nutrient  content,  and  relative  digestibility  by  ursids 
of  whitebark  pine  seeds;  compiled  from  Craighead  and 
others  (1982),  Kendall  (1983),  Lanner  and  Gilbert  (these 
proceedings),  Mealey  (1975),  Pritchard  and  Bobbins 
(1990),  Weaver  and  Forcella  (1986),  and  unpublished 
Interagency  Grizzly  Bear  Study  Team  data 


Mean 

Range  of 

Range 

among 

standard 

Parameters 

of  means 

studies 

deviations 

Seeds  per  cone 

75-88 

80 

26-28 

Dry  seed  weight  (g) 

0.07-0.10 

0.09 

0.01-0.02 

Percent  protein  content^ 

11.9-14.2 

12.8 

Percent  ether  extract^ 

21 .8-30.4 

27.1 

Percent  nitrogen-free  extract' 

12.4-27.2 

19.8 

Percent  crude  fiber' 

34.8 

(34.8) 

Percent  total  dietary  fiber' ^ 

40.3 

(40.3) 

Percent  digestible  energy 

35O.I/M8.7 

49.4 

31.4 

'Based  on  whole  seed  dry  weight. 
^Calculated  for  pinyon  pine  (Pinus  edulis)  seeds. 
^Calculated  based  on  laboratory  feeding  trials  (Pritchard  and  Robbins 
1990). 

^Calculated  from  differential  between  field  samples  of  fresh  whole  material 
and  feces  (Mealey  1975). 


212 


Figure  1 — Distribution  of  whitebark  pine  and  use  of  whitebark  pine  seeds  by  bears 
in  North  America.  Sparser  Canadian  distribution  of  whitebark  pine  is  indicated  by 
dotted  lines.  Three  different  levels  of  pine  seed  use  (<1  percent,  1-10  percent,  and 
>10  percent  frequency  in  bear  feces)  are  also  indicated. 


213 


with  continental  climates  compared  to  areas  with  mari- 
time climates  (2.7  percent)  or  areas  in  Canada  where 
whitebark  pine  was  sparse  (2.8  percent)  (df  =  2,  G  =  987.3, 
P<  0.001). 

These  differences  are  logically  attributable  to  several 
factors.  In  areas  where  whitebark  pine  is  scarce,  the 
bears  may  not  consider  it  to  be  a  food,  or  not  choose  to  use 
its  seeds  if  they  are  disjunct  from  other  higher  value  feed- 
ing opportunities.  Areas  influenced  by  maritime  climates 
are  characterized  by  an  abundance  of  berry-producing 
shrubs,  typically  at  elevations  much  lower  than  whitebark 
pine  (Jonkel  and  Cowan  1971;  Poelker  and  Hartwell  1973; 
Servheen  1983;  Tisch  1961;  and  others).  These  berries 
are  probably  more  valuable  than  whitebark  pine  seeds, 
and  so  local  bears  would  logically  tend  to  use  berry-rich 
habitats  to  the  exclusion  of  whitebark  pine  habitats  (Tisch 
1961).  Conversely,  in  areas  with  continental  climates  ber- 
ries are  typically  fewer  and  whitebark  pine  more  abun- 
dant (Aune  cuid  Kasworm  1989;  Mattson  and  others 
1991).  We  hypothesize  that  for  this  reason  bears  would 
logically  focus  more  on  the  consumption  of  whitebark  pine 
seeds.  Recent  heavy  whitebark  pine  mortality  in  mari- 
time areas,  caused  by  white  pine  blister  rust,  has  also  ar- 
guably had  negative  effects  on  bear  use  of  pine  seeds  in 
these  areas  (Kendall  and  Arno  1990;  Mattson  and  Jonkel 
1990). 

Because  the  large  majority  of  whitebark  pine  seed  use 
by  bears  has  occurred  in  interior  ecosystems,  our  subse- 
quent discussion  will  focus  on  these  areas;  and  because 
most  research  concerning  bear  use  of  pine  seeds  comes 
from  the  Yellowstone  ecosystem,  most  of  the  results  pre- 
sented here  are  from  this  study  area. 


APR  MAY  JUN  JUL  AUG  SEP  OCT 
n  =      5  10  14  14  14  11  6 


Figure  2 — Mean  percent  volume  of  pine 
seeds  in  Yellowstone  grizzly  bear  feces  by 
month,  1977-90.  Maximum  recorded  volume 
is  indicated  by  dashed  lines  and  mean  per- 
cent volume  of  pine  seeds  in  feces  of  occur- 
rence is  shown  in  the  inset. 


Temporal  Variation  in  Use 

Seasonal  Variation — Because  whitebark  pine  seeds 
are  durable,  bears  use  not  only  seeds  contemporary  to 
their  maturation,  but  also  seeds  that  survive  for  up  to 
1  year  after  initial  availability  of  a  large  crop  (Kendall 
1983;  Mattson  and  Jonkel  1990;  Mattson  and  others  1993). 
Current-year  use  usually  begins  after  mid-August  with 
onset  of  caching  by  red  squirrels  (Hutchins  and  Lanner 
1982),  but  can  vary  from  late  July  to  early  September, 
apparently  depending  on  seed  maturation  (Mattson  and 
others  1993).  Use  of  overwintered  seeds  does  not  usually 
begin  imtil  after  mid-May  (Mattson  and  Jonkel  1990),  but 
can  begin  as  early  as  the  first  week  of  April  and  involve 
excavation  of  cones  out  of  squirrel  middens,  through  snow 
0.3-1.2  m  deep  (Kendall  1983). 

Use  of  both  current-year's  and  overwintered  seeds  by 
bears  results  in  a  primary  September-October  peak  and 
a  secondary  June  peak  in  frequency  of  pine  seeds  in  bear 
feces,  when  averaged  over  a  number  of  years  (fig.  2). 
With  the  exception  of  July,  bear  feces  tend  to  consist 
almost  exclusively  of  pine  seed  remains  (>80  percent) 
when  bears  are  in  the  whitebark  pine  zone  feeding  on 
pine  seeds  (fig.  2). 

Annual  Variation — Use  of  pine  seeds  by  bears  varies 
considerably  among  years.  In  the  Yellowstone  area,  pine 
seeds  have  dominated  the  diet  during  some  years  (for  ex- 
ample 1979)  and  been  virtually  absent  other  years  (for 
example  1977,  1982,  and  1988)  (Mattson  and  others  1991, 
1992).  This  variation  in  use  is  related  to  seed  crop  size, 
with  heavy  use  occurring  only  when  crops  average  >  13-23 
cones  per  tree  (from  counts  on  permanent  transects),  ap- 
proaching an  asymptote  of  use  (75  percent  frequency  in 
feces)  at  crop  sizes  averaging  38  cones  per  tree.  Use  of 
overwintered  pine  seeds  during  the  following  June-July  is 
similarly  related  to  crop  size,  but  with  inflection  points  of 
20-29  seeds  and  an  asymptote  of  38  percent  fi-equency  in 
feces  (fig.  3).  This  relationship  between  bear  use  and  crop 
size  is  acutely  sigmoidal,  suggesting  the  existence  of  a  use 
threshold  that  is  possibly  related  to  sattiration  of  demand 
by  competing  birds  and  rodents,  nonlinear  increases  in 
foraging  efficiency,  or  both.  The  higher  thresholds  and 
lower  asymptote  associated  with  use  of  overwintered 
seeds  is  attributable  to  depletion  of  the  crop  diiring  the 
previous  fall  and  winter. 

Varying  inflection  points  in  the  hypothesized  relation- 
ship between  bear  use  and  seed  crop  size  reflect  an  appar- 
ent trend  over  the  years  1980-91  toward  lower  thresholds 
to  substantial  bear  use  (fig.  3).  The  years  1980  and  1981 
are  further  distinguished  by  the  relative  insensitivity  of 
bear  use  to  crop  size  compared  to  later  years.  Interest- 
ingly, asymptotes  associated  with  maximimi  potential 
bear  use  have  not  varied  significantly  diiring  this  same 
time  period.  At  the  very  least,  this  variance  suggests 
that  the  relative  intensity  with  which  bears  use  pine 
seeds  in  a  given  study  area  can  change.  The  causes  of 
these  apparent  changes  can  only  be  speculated,  including: 
increasingly  efficient  use  of  comparable-sized  crops  by  the 
Yellowstone  bear  population;  and  an  increasing  willing- 
ness to  use  seeds  imder  less  optimal  conditions,  attribut- 
able either  to  increasing  bear  densities  or  decreasing  over- 
all carrying  capacity. 


214 


AUGUST-OCTOBER  USE,  CONTEMPORARY  YEAR 


JUNE-JULY  USE,  FOLLOWING  YEAR 


100 


75 


50 


25  — 


-[((CONES/60)-1)/I]'^     r^  =  0  99 
%  FREQ  =  38.0  *  e  f  =  353  3 


0 


B 


10        20        30  40 

MEAN  NUMBER  OF  CONES  PER  TREE 


50 


Figure  3 — A,  relationship  between  frequency 
of  whitebark  pine  seeds  in  August-October 
Yellowstone  grizzly  bear  feces  and  mean  num- 
ber of  cones  counted  per  tree  on  permanent 
transects  during  the  contemporary  year,  1980- 
91 .  B,  relationship  between  frequency  of  pine 
seeds  in  June-July  feces  and  mean  number  of 
cones  counted  the  previous  fall,  1 980-91 . 


Interestingly,  in  a  part  of  Montana  characterized  by 
a  more  maritime  climate  and  relatively  little  bear  use  of 
pine  seeds,  Tisch  (1961)  counted  an  average  of  five  and 
16  cones  per  tree  during  1959  and  1960,  respectively. 
Both  of  these  averages  were  less  than  the  approximately 
20  cones  per  tree  associated  with  the  onset  of  heavy  pine 
seed  use  in  the  Yellowstone  area. 

Habitats  Associated  With  Use 

Elevation  and  Aspect — As  would  be  expected,  bear 
use  of  whitebark  pine  seeds  is  restricted  to  elevations 
above  the  lower  limits  of  whitebark  pine  distribution. 
This  lower  elevational  limit  varies  with  latitude  and 
longitude,  from  about  2,400  m  in  the  Yellowstone  area 


(44.5°N.).  to  1,890  m  along  the  East  Front  of  the  Montana 
Rockies  (47.5° N.)  (Mattson  and  Jonkel  1990). 

However,  within  this  constraint  bear  use  varies  consid- 
erably among  aspects  and  elevations,  depending  on  the 
year.  We  found  that  diiring  three  successive  large  seed 
crops  in  the  Yellowstone  area,  bear  use  was  nonrepetitive 
among  broad  aspectual  and  elevational  categories;  during 
1985-86  heaviest  on  east  aspects  at  mid-elevations,  during 
1987  heaviest  on  west  and  north  aspects  at  low-  to  mid- 
elevations,  and  dxuing  1989-90  heaviest  on  east  and  south 
aspects  at  high  elevations  (Mattson  and  others  1993).  We 
also  found  differences  in  elevational  distribution  of  pine 
seed  use  between  two  successive  crops  at  a  local  scale, 
in  the  Cooke  City  area  northeast  of  Yellowstone  National 
Park,  with  use  concentrated  at  elevations  >2,870  m 
dxoring  1990  and  at  elevations  <2,720  m  during  1991 
(Reinhart  and  Mattson  1992).  This  geographic  variation 
was  inconclusively  related  to  seed  production,  but  demon- 
strated the  potential  for  considerable  differences  in  habi- 
tat distribution  of  pine  seed  use  over  a  6-  to  7-year  period. 

Habitat  Tjrpes — Ecosystems  in  the  Western  United 
States  have  been  described  in  terms  of  ecological  land- 
scape units  called  habitat  types,  that  are  further  aggre- 
gated as  series  and  split  into  phases  (Pfister  and  others 
1977;  Steele  and  others  1983).  In  the  Yellowstone  area 
most  bear  use  of  pine  seeds  has  occurred  with  varying  in- 
tensity in  14  different  habitat  types.  In  accord  with  varia- 
tion in  distribution  of  pine  seed  use  among  aspects  and  el- 
evations, distribution  of  pine  seed  use  by  bears  has  also 
varied  among  habitat  types  (Mattson  and  others  1993). 
Even  so,  the  Abies  lasiocarpa  /Vaccinium  scoparium- 
Pinus  albicaulis  phase,  and  the  Abies  lasiocarpa  / 
Calamagrostis  canadensis  and  Abies  lasiocarpa  I 
Thalictrum  occidentale  habitat  types  have  been  used 
intensively  during  all  years  when  pine  seeds  are  abtm- 
dant  (Mattson  and  Jonkel  1990).  Conversely,  use  of  habi- 
tat types  at  lowest  and  highest  elevations  of  the  white- 
bark pine  zone  has  been  most  varied.  Habitats  in  interior 
ecosystems  where  whitebark  pine  is  a  chmax  overstory 
species  have  been  only  rarely  used  by  bears  for  foraging 
on  pine  seeds  (Mattson  and  Jonkel  1990).  However,  even 
in  these  habitats  heavy  use  can  occur,  as  diiring  1989-90 
in  the  Yellowstone  area  (Mattson  and  others  1993). 

Distribution  of  pine  seed  use  among  habitat  types 
was  surprisingly  consistent  between  study  areas  in  the 
Yellowstone  ecosystem  and  along  the  East  Front  of  the 
Montana  Rocky  Mountains  (Mattson  and  Jonkel  1990). 
This  suggests  that  patterns  of  habitat  use  associated  with 
bear  consumption  of  pine  seeds  are  consistent  among 
areas  vvdth  continental  climates,  and  that  more  detailed 
results  pertaining  to  habitat  use  from  the  Yellowstone 
ecosystem  are  extrapolable  over  interior  regions  of  the 
United  States.  Because  very  Uttle  information  is  avail- 
able concerning  bear  use  of  pine  seeds  in  more  maritime 
ecosystems,  we  can  make  no  inferences  about  patterns  of 
use  in  these  areas  other  than  it  being  restricted  to  sites 
with  mature  cone-producing  whitebark  pine. 

Red  Squirrel  and  Whitebark  Pine  Abundance — 

Red  squirrel  density  is  probably  the  most  important  proxi- 
mal factor  affecting  bear  use  of  pine  seeds  in  interior  eco- 
systems. The  importance  of  red  squirrels  to  especially 


215 


grizzly  bear  foraging  has  been  observed  by  Kendall  (1983), 
Mattson  and  Jonkel  (1990),  and  Schallenberg  and  Jonkel 
(1980).  Craighead  and  others  (1982),  Graber  and  White 
(1983),  Murie  (1944),  and  Raine  and  Kansas  (1990)  also 
mentioned  rodent  or  squirrel  caches  as  potential  sources 
of  pine  seeds  for  bears,  but  were  more  equivocal  about  the 
overall  importance  of  these  intermediaries.  Interestingly, 
neither  Craighead  and  Craighead  (1972)  nor  Mealey 
(1975)  mention  that  grizzly  bears  raided  red  squirrel  food 
caches  during  their  studies  in  Yellowstone.  The  disparity 
between  these  and  later  observations  in  the  Yellowstone 
area  is  not  easily  reconciled,  and  suggests  that  differences 
could  be  at  least  partly  attributable  to  variation  in  bear 
behavior  and  pine  seed  availability  described  earlier. 

Most  (86  percent)  of  232  sites  where  grizzly  bears  were 
known  to  have  used  pine  seeds  in  the  Yellowstone  area, 
1986-91,  involved  bears  excavating  red  squirrel  cone 
caches.  All  but  five  of  the  other  32  instances  where  griz- 
zly acquired  pine  seeds  by  scavenging  fallen  cones  oc- 
curred during  1989-90  when  much  of  the  cone  production 
occurred  at  elevations  above  the  normal  range  of  red 
squirrels  (Mattson  and  others  1993;  Reinhart  and 
Mattson  1990).  Conversely,  51  percent  of  53  sites  where 
black  and  grizzly  bears  used  pine  seeds  near  Cooke  City, 
MT,  during  1990-91  were  characterized  by  scavenging  of 
fallen  cones  (Reinhart  and  Mattson  1992).  This  observa- 
tion suggests  that  scavenging  of  fallen  cones  can  be  locally 
important.  However,  all  26  instances  when  bears  scav- 
enged fallen  cones  in  this  area  occurred  during  1990,  in 
conjunction  with  use  of  seeds  overwintered  from  the  im- 
usually  high-elevation  crop  of  1989. 

Scavenging  apparently  predominates  over  excavation 
of  squirrel  caches  in  environments  where  squirrels  are 
rare.  Red  squirrel  densities  in  the  whitebark  pine  zone 
are  strongly  related  to  a  site  favorability  index  (r^  =  0.79), 
with  squirrel  densities  increasing  as  exposm*e  to  prevail- 
ing winds  and  elevation  decreases  (Reinhart  and  Mattson 
1990).  Given  that  basal  area  and  exclusivity  of  whitebark 
pine  increase  with  wind  exposure  and  elevation  (Mattson 
and  Reinhart  1990),  there  is  an  inherently  negative  rela- 
tionship between  dominance  of  whitebark  pine  and  den- 
sity of  red  squirrels  within  the  whitebark  pine  zone.  Thus, 
scavenging  predominates  over  midden  excavation  where 
whitebark  pine  comprises  >75  percent  of  the  overstory 
(four  categories  of  whitebark  pine  percentage;  df  =  3, 
G  =  69.7,  P  <  0.001),  and  at  elevations  >2,870  m  (four 
elevation  categories;  df  =  3,  G  =  42.4,  P  <  0.001);  where 
red  squirrels  are  rare. 

Bear  use  of  pine  seeds  is  clearly  the  result  of  opportuni- 
ties implicit  to  the  varying  abundance  of  red  squirrels  and 
whitebark  pine.  However,  there  is  reason  to  hypothesize 
that  squirrel  densities  impose  greater  constraints  on  bear 
use  of  pine  seeds  than  does  abundance  of  whitebark  pine. 
We  analyzed  data  from  1987  to  examine  this  hypothesis, 
looking  at  the  relationship  between  unit  area  volume  of 
excavations  for  pine  seeds  by  Yellowstone  grizzly  bears 
and  red  squirrel  density  and  whitebark  pine  basal  area 
(fig.  4).  At  any  given  squirrel  density,  bear  use  abruptly 
increased  and  approached  maximum  levels  at  relatively 
low  basal  areas  of  whitebark  pine,  between  approximately 
3  and  8  m^/ha.  Conversely,  the  relationship  between  bear 
activity  and  red  squirrel  densities  was  more  nearly  linear. 


BEAR  USE  =  22.6  •  (e 
^Opr 

<  I 


-[((SD/3.68)-1)/0.382]' 


)*(e 


-[((BA/254)-1)/0.983l* 


< 
ui 

QC 
< 


20 


< 

(0 

< 
m 

UJ 

z 

a. 

bi. 
a. 
< 
m 

LU 
X 

5 


10 


1.0  2.0  3.0  ^.0 

SQUIRREL  DENSITY  (N/KM)  (SD) 

Figure  3 — The  relationship  between  intensity  of 
grizzly  bear  feeding  on  pine  seeds,  as  indicated  by 
density  of  excavations  (m^/km)  and  squirrel  density 
and  whitebark  pine  basal  area,  Yellowstone  area, 
1987. 


with  excavations  gradually  increasing  to  a  maximum  that 
corresponded  to  maximum  squirrel  densities  observed  in 
the  whitebark  pine  zone.  Basically,  whitebark  pine  basal 
area  seemed  to  be  associated  primarily  with  the  probabil- 
ity that  a  given  midden  would  be  excavated,  sometimes 
repeatedly,  by  bears  (Mattson  and  Reinhart  1987),  with 
the  greatest  effects  at  very  low  whitebark  pine  abun- 
dance. Once  this  low  threshold  of  whitebark  pine  abun- 
dance was  exceeded  (>  about  4.3  m^/ha),  red  squirrel  den- 
sities appeared  to  limit  bear  use  of  pine  seeds  during  most 
years. 

Timber  Stand  Age — Bear  use  of  whitebark  pine  seeds 
in  the  Yellowstone  area  has  occurred  almost  exclusively  in 
mature  stands.  According  to  a  successional  classification 
of  timber  types  by  Despain  (1990),  all  bear  use  of  pine 
seeds  from  squirrel  middens  and  most  bear  use  of  scav- 
enged seeds  has  occurred  in  "mature"  cover  types,  typi- 
cally designated  LP2,  LP3,  SF,  WB2,  WB3,  WB,  or  DF3. 
Results  from  a  use-availability  analysis  of  cover  types  in 
Yellowstone  National  Park  corroborate  the  affinity  of  griz- 
zly bears  for  later  successional  cover  types  when  using 
pine  seeds  (Knight  and  others  1984).  This  analysis  showed 
that  the  early  successional  WBl  cover  type  was  consis- 
tently imder-used  or  used  without  preference  during  all 
seasons,  whereas  mature  WB  cover  types  were  consistently 
used  with  disproportionally  greater  intensity  than  ex- 
pected, especially  dining  the  fall,  primarily  for  the  con- 
sumption of  pine  seeds. 


216 


Based  on  ages  given  by  Despain  (1990)  cover  types  asso- 
ciated with  pine  seed  use  are  rarely  less  than  100  years 
old.  Ages  of  160  stands  in  two  intensive  Yellowstone 
study  areas  located  in  the  whitebark  pine  zone  (Mattson 
and  Reinhart  1987)  also  revealed  that  none  of  these  ma- 
ture  cover  types  was  <125  years  old. 

These  minimum  stand  ages  correspond  with  use  ex- 
pected by  levels  of  cone  production.  Weaver  and  others 
(1990)  observed  that  whitebark  pine  cone  production  was 
relatively  low  in  stands  <100  years  old  and  reached  peak 
levels  soon  thereafter.  Morgan  and  Biinting  (in  press) 
also  observed  that  younger  WBO  and  WBl  stands  pro- 
duced few  cones  and  had  few  excellent  crops.  Both  of 
these  authors  observed  that  cone  production  persisted  at 
least  at  moderate  levels  in  much  older  stands,  up  to  300  to 
400  years  old.  These  observations  correspond  well  with 
the  resvilts  of  more  exhaustive  studies  of  the  closely  re- 
lated Siberian  stone  pine  (Pinus  sibirica)  (Axelrod  1986) 
by  Russian  researchers  (Mattson  and  Jonkel  1990). 

These  relationships  of  use  to  stand  age  have  several  im- 
plications. First,  potential  productivity  of  whitebark  pine 
sites,  at  least  for  bears,  is  not  realized  until  after  approxi- 
mately 100  years,  and  can  persist  for  200  to  300  years 
thereafter.  Thus,  in  contrast  to  many  other  temperate 
ecosystems  where  production  of  bear  foods  is  concentrated 
on  berry-producing  shrubs  in  early  successional  stages, 
production  of  bear  foods  in  the  whitebark  pine  zone  is  con- 
centrated in  much  older  mid-  to  late-successional  stands 
(Mattson  1990).  Second,  given  these  time  frames,  long- 
term  productivity  of  the  whitebark  pine  zone  in  interior 
ecosystems  would  be  best  assiired  by  disturbance  regimes 
that  replaced  approximately  3  percent  of  mature  acreage 
per  decade. 

EFFECTS  ON  BEAR  DEMOGRAPHY 
Fecundity 

Whitebark  pine  seeds  are  an  important  so\irce  of  energy 
for  bears  that  live  in  interior  ecosystems  of  the  United 
States.  Craighead  and  others  (1982)  estimated  that  pine 
seeds  were  the  most  important  soxirce  of  energy  for  grizzly 
bears  in  their  study  area  located  east  of  the  Continental 
Divide  in  the  Rocky  Mountains  of  Montana.  Preliminary 
estimates  by  Servheen  and  others  (1986)  suggested  that 
Yellowstone  grizzly  bears  derived  about  23  percent  of 
their  net  digested  energy  from  whitebark  pine  seeds, 
second  only  to  the  approximately  30  percent  that  they 
derived  from  ungulate  meat.  In  addition,  preliminary 
analysis  of  Yellowstone  data  suggests  that  female  grizzly 
bears  make  proportionately  much  greater  use  of  pine 
seeds  compared  to  males.  Thus  there  is  a  good  basis  for 
hypothesizing  that  availability  and  use  of  pine  seeds 
affects  the  fec;mdity  of  female  bears  residing  in  these 
ecosystems. 

Further  preliminary  analyses  of  data  from  the  Yellow- 
stone ecosystem  offer  support  for  this  hypothesis.  Female 
grizzly  bears  that  frequently  used  whitebark  pine  seeds 
appeared  to  exhibit  substantially  higher  reproductive 
rates  and  reproduce  at  a  significantly  earlier  age  than 
females  who  consumed  few  pine  seeds.  Much  of  the  differ- 
ence in  reproductive  rate,  as  measured  by  both  production 


and  survival  of  offspring  to  weaning,  was  attributable  to 
low  survivorship  of  cubs  dependent  on  dams  that  made 
httle  lose  of  pine  seeds. 

Mortality  and  Movements 

Whitebark  pine  seed  availability  has  even  greater  ef- 
fects on  svirvivorship  of  subadult  males  and  adult  females 
than  on  fecimdity  of  reproducing  females.  During  years 
when  pine  seeds  are  used  very  little,  conflicts  with  hu- 
mans escalate  dramatically  in  the  Yellowstone  area. 
Management-related  trappings  of  grizzly  bears  instituted 
to  alleviate  these  conflicts  are,  on  average,  six  times  as 
nimierous  during  nonuse  compared  to  use  years  (Mattson 
and  others  1992).  As  a  direct  result  of  this  increased  con- 
flict, an  average  of  two  times  as  many  adult  females  and 
three  times  as  many  subadult  males  die  during  years 
when  pine  seeds  are  not  available  (Mattson  and  others 
1992). 

This  increased  mortality  is  not  attributable  to  increased 
movement  of  htunans  smiong  bears  or  to  starvation,  but 
rather  to  increased  activity  of  bears  around  hvunan  facili- 
ties (Mattson  and  others  1992).  Whitebark  pine  in  the 
Yellowstone  and  Rocky  Movmtain  East  Front  ecosystems 
tends  to  occm*  in  areas  remote  from  human  facilities  (Aune 
and  Kasworm  1989;  Mattson  and  others  1992).  Thus, 
when  bears  are  intensively  using  pine  seeds  they  are  typi- 
cally far  removed  from  himaans  and  hmnan-bear  conflict 
is  correspondingly  low.  Conversely,  when  pine  seeds  are 
not  available  bears  seek  out  alternate  foods  that  are  often 
closely  associated  with  himian  facilities.  Historically 
these  alternate  foods  were  often  of  human  origin  (Knight 
and  others  1988),  but  more  recently  these  alternate  foods 
have  been  natives  that  tend  to  occupy  lower  elevation 
habitats  also  occupied  by  humans  (Mattson  and  Knight 
1989).  In  addition,  the  greater  movements  of  Yellowstone 
grizzly  bears  diiring  years  or  seasons  when  pine  seeds  are 
not  available  (Blanchard  and  Knight  1991;  Haroldson  and 
Mattson  1985)  very  likely  exacerbate  this  unfavorable 
situation  and  increase  conflicts  with  himaans. 

MANAGEMENT  IMPLICATIONS 

Because  whitebark  pine  seeds  are  apparently  so  impor- 
tant to  bears  in  interior  ecosystems  of  the  United  States, 
whitebark  pine  deserves  priority  in  the  management  of 
bear  habitat.  As  indicated  earlier,  both  temporal  and  geo- 
graphic variation  in  the  availability  of  whitebark  pine 
seeds  can  have  major  implications  on  the  survival  and  fe- 
cimdity of  bears  in  these  ecosystems.  Thus,  long-term 
variations  in  availability  of  pine  seeds  attributable  to  the 
productivity  and  abundance  of  whitebark  pines  warrants 
the  concern  and  attention  of  managers. 

As  pointed  out  previously,  stands  with  relatively  little 
whitebark  pine  can  be  of  considerable  value  to  bears  as 
long  as  they  contain  high  densities  of  red  squirrels.  Thus, 
management  of  bear  habitat  for  availability  of  pine  seeds 
implies  managment  to  meet  red  squirrel  habitat  require- 
ments. In  the  whitebark  pine  zone  this  means  simulta- 
neously maximizing  stand  basal  area,  overstory  species 
diversity,  and  components  of  whitebark  pine  and  Douglas- 
fir  (Pseudotsuga  menziesii)  (Reinhart  and  Mattson  1990). 


217 


This  kind  of  favorable  condition  often  coincides  with  com- 
mercial forests  occupying  lower  elevations  of  the  white- 
bark  pine  zone;  and  maintenance  of  these  conditions  po- 
tentially conflicts  with  other  management  objectives 
related  to  the  extraction  of  wood  products.  In  areas  where 
bear  management  is  a  high  priority,  timber  harvest  cycles 
should  therefore  employ  roughly  300-year  stand  rotations 
in  the  whitebark  pine  zone,  with  landscape-wide  harvest 
levels  at  approximately  3  percent  per  decade. 

Global  climate  change  has  potentially  major  implica- 
tions to  the  abundance  of  whitebark  pine  in  interior  eco- 
systems and  consequently  to  the  prospects  of  associated 
bear  populations.  Romme  and  Turner  (1991)  projected 
the  fates  of  major  life  zones  in  the  Yellowstone  area  under 
three  different  climatic  scenarios  likely  to  be  associated 
with  doubled        content  in  the  atmosphere.  Under  all 
three  scenarios  they  projected  a  decline  of  the  whitebark 
pine  zone  to  <10  percent  of  its  current  extent.  Without 
wildfires,  loss  of  whitebark  pine  may  be  prolonged  due  to 
the  long  lifespan  of  the  species.  However,  increasingly 
frequent  fires,  especially  under  a  drier  scenario,  could 
hasten  the  demise  of  whitebark  pine  by  eliminating  it 
from  sites  where  it  coiild  not  reestablish  (Romme  and 
Turner  1991).  Thus,  with  global  climate  warming  a  real 
possibility,  managers  should  be  cautious  about  harvesting 
any  whitebark  pine,  especially  at  lower  elevations.  In  ad- 
dition, judicious  control  of  natural  fires  may  be  desirable 
to  favor  the  persistence  of  existing  mature  whitebark 
pine. 

The  potential  and  demonstrated  negative  impacts  of 
global  climate  change  and  white  pine  blister  rust  on  an 
important  source  of  bear  food  like  whitebark  pine  point  up 
the  legitimate  need  for  concern  about  long-term  changes 
in  bear  habitats.  This  consideration  is  especially  relevant 
to  the  management  of  insular  bear  populations,  exempli- 
fied by  Yellowstone's  grizzly  bears  (Mattson  and  Reid 
1991).  In  situations  where  occupied  bear  habitat  is  not 
only  isolated  but  also  restricted,  bear  populations  are  es- 
pecially vulnerable  to  changes  in  carrying  capacity  be- 
cause of  limited  options  to  shift  or  expand  their  ranges  or 
benefit  from  the  infliix  of  bears  from  other  nearby  or  con- 
tiguous populations.  Thus,  uncertainty  over  the  long- 
term  prospects  of  whitebark  pine  should  be  a  major  con- 
sideration in  management  decisions,  especially  those  that 
might  otherwise  reduce  habitat  capability  under  the  as- 
sumption of  long-term  habitat  stability. 

Finally,  managers  should  be  concerned  about  the  place- 
ment of  additional  human  facilities  anywhere  in  the 
whitebark  pine  zone.  Currently,  most  of  the  whitebark 
pine  zone  serves  as  a  refuge  for  bears  from  conflict  with 
humans,  at  least  when  pine  seeds  are  available.  This  ref- 
uge effect  will  be  diminished  with  every  additional  human 
intrusion.  Wherever  bear  management  is  a  priority,  addi- 
tional human  facilities  should  not  be  constructed  in  the 
whitebark  pine  zone  and  existing  facilities  should  be  re- 
moved if  at  all  possible. 

REFERENCES 

Aune,  Keith;  Kasworm,  Wayne.  1989.  Final  report  East 
Front  grizzly  studies.  Helena,  MT:  Montana  Depart- 
ment of  Fish,  Wildlife,  and  Parks.  332  p. 


Axelrod,  Daniel  I.  1986.  Cenozoic  history  of  some  western 
American  pines.  Annals  of  the  Missouri  Botanical  Gar- 
dens. 73:  565-641. 

Blanchard,  Bonnie  M.;  Knight,  Richard  R.  1991.  Move- 
ments of  Yellowstone  grizzly  bears,  1975-87.  Biological 
Conservation.  58:  41-67. 

Cahalane,  Victor  H.  1947.  Mammals  of  North  America. 
New  York:  Macmillan.  682  p. 

Craighead,  Frank  C,  Jr.;  Craighead,  John  J.  1972.  Griz- 
zly bear  prehibernation  and  denning  activities  as  deter- 
mined by  radio-tracking.  Wildlife  Monograph  32.  35  p. 

Craighead,  John  J.;  Sumner,  J.  S.;  Scaggs,  G.  B.  1982.  A 
definitive  system  for  analysis  of  grizzly  bear  habitat 
and  other  wilderness  resources.  Wildlife-Wildlands 
Institute  Monograph  1.  Missoula,  MT:  University  of 
Montana,  U  of  M  Foundation.  279  p. 

Despain,  Donald  G.  1990.  Yellowstone  vegetation, 
Boiilder,  CO:  Roberts  Reinhart.  239  p. 

Graber,  David  M.;  White,  Marshall.  1983.  Black  bear  food 
habits  in  Yosemite  National  Park.  In:  International 
Conference  on  Bear  Research  and  Management:  Pro- 
ceedings: 5:  1-10. 

Hadley,  N.  F.  1985.  The  adaptive  role  of  lipids  in  biologi- 
cal systems.  New  York:  John  Wiley  and  Sons.  319  p. 

Haroldson,  Mark;  Mattson,  David  J.  1985.  Response  of 
grizzly  bears  to  backcountry  himian  use  in  Yellowstone 
National  Park.  Bozeman,  MT:  U.S.  Department  of  the 
Interior,  National  Park  Service,  Interagency  Grizzly 
Bear  Study  Team.  38  p. 

Jonkel,  Charles  J.;  Cowan,  Ian  McT.  1971.  The  black  bear 
in  the  spruce-fir  forest  of  northwestern  Montana.  Wild- 
life Monograph.  27.  55  p. 

Kendall,  Katherine  C.  1983.  Use  of  pine  nuts  by  black  and 
grizzly  bears  in  the  Yellowstone  area.  In:  International 
Conference  on  Bear  Research  and  Management:  Pro- 
ceedings: 5:  166-173. 

Kendall,  Katherine  C;  Arno,  Stephen  F.  1990.  Whitebark 
pine — an  important  but  endangered  wildlife  resource. 
In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
264-273. 

Knight,  R.  R.;  Blanchard,  B.  M.;  Eberhardt,  L.  L.  1988. 
Mortality  patterns  and  population  sinks  for  Yellowstone 
grizzly  bears,  1973-1985.  Wildlife  Society  Bulletin.  16: 
121-125. 

Knight,  Richard  R.;  Mattson,  David  J.;  Blanchard, 
Bonnie  M.  1984.  Movements  and  habitat  use  of  the 
Yellowstone  grizzly  bear.  Bozeman,  MT:  U.S.  Depart- 
ment of  the  Interior,  National  Park  Service,  Interagency 
Grizzly  Bear  Study  Team.  177  p. 

Kurten,  Bjorn;  Anderson,  Elaine.  1980.  Pleistocene  mam- 
mals of  North  America.  New  York:  Columbia  University 
Press.  512  p. 

Lanner,  Ronald  M.  1990.  Biology,  taxonomy,  evolution, 
and  geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989  March 


218 


29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermountain  Research  Station:  14-24. 

Mattson,  David  J.  1990.  Human  impacts  on  bear  habitat 
use.  In:  International  Conference  on  Bear  Research  and 
Management:  Proceedings:  8:  33-56. 

Mattson,  David  J.;  Blanchard,  Bonnie  M.;  Knight,  Richard 
R.  1991.  Food  habits  of  Yellowstone  grizzly  bears,  1977- 
1987.  Canadian  Journal  of  Zoology.  69:  1619-1629. 

Mattson,  David  J.;  Blanchard,  Bonnie  M.;  Knight,  Richard 
R.  1992.  Yellowstone  grizzly  bear  mortality,  hvmaan  ha- 
bituation, and  whitebark  pine  seed  crops.  Journal  of 
Wildlife  Management.  56(3):  432-442. 

Mattson,  David  J.;  Jonkel,  Charles.  1990.  Stone  pines  and 
bears.  In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine 
ecosystems:  ecology  and  management  of  a  high-moun- 
tain resource;  1989  March  29-31;  Bozeman,  MT.  Gen. 
Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermountain  Research 
Station:  223-236. 

Mattson,  David  J.;  Knight,  Richard  R.  1989.  Evaluation  of 
grizzly  bear  habitat  using  habitat  type  and  cover  type 
classifications.  In:  Ferguson,  Dennis  E.;  Morgan, 
Penelope;  Johnson,  Frederic  D.,  comps.  Proceedings — 
land  classifications  based  on  vegetation:  applications  for 
resource  management;  1987  November  17-19;  Moscow, 
ID.  Gen.  Tech.  Rep.  INT-257.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain 
Research  Station:  135-143. 

Mattson,  David  J.;  Reid,  Matthew  M.  1991.  Conservation 
of  the  Yellowstone  grizzly  bear.  Conservation  Biology. 
5:  364-372. 

Mattson,  David  J.;  Reinhart,  Daniel  P.  1987.  Grizzly  bear, 
red  squirrels,  and  whitebark  pine:  third  year  progress 
report.  In:  Knight,  Richard  R.;  Blanchard,  Bonnie  M.; 
Mattson,  David  J.  Yellowstone  grizzly  bear  investiga- 
tions: report  of  the  Interagency  Study  Team,  1986. 
Bozeman,  MT:  U.S.  Department  of  the  Interior, 
National  Park  Service,  Interagency  Grizzly  Bear  Study 
Team:  36-45. 

Mattson,  David  J.;  Reinhart,  Daniel  P.  1990.  Whitebark 
pine  on  the  Mount  Washburn  massiff,  Yellowstone  Na- 
tional Park.  In:  Schmidt,  Wyman  C;  McDonald,  Kathy 
J.,  comps.  Proceedings — symposium  on  whitebark  pine 
ecosystems:  ecology  and  management  of  a  high-moun- 
tain resource;  1989  March  29-31;  Bozeman,  MT.  Gen. 
Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermountain  Research 
Station:  106-117. 

Mattson,  David  J.;  Reinhart,  Daniel  P.;  Blanchard, 
Bonnie  M.  1993.  Variation  in  production  and  bear  use  of 
whitebark  pine  seeds  in  the  Yellowstone  area.  In:  Pro- 
ceedings— conference  on  plants  and  their  environments; 
1991  September;  Mammoth,  WY.  Denver,  CO:  U.S. 
Department  of  the  Interior,  National  Park  Service, 
Regional  Office.  [In  press]. 

Mealey,  Stephen  Patrick.  1975.  The  natural  food  habits  of 
free  ranging  grizzly  bears  in  Yellowstone  National 
Park,  1973-1974.  Bozeman,  MT:  Montana  State  Univer- 
sity. 158  p.  Thesis. 


Murie,  O.  J.  1944.  Progress  report  on  the  Yellowstone 
bear  study.  Mammoth,  WY:  U.S.  Department  of  the 
Interior,  National  Park  Service,  Yellowstone  National 
Park.  13  p. 

Pfister,  Robert  D.;  Kovalchik,  Bernard  L.;  Arno,  Stephen  F.; 
Presby,  Richard  C.  1977.  Forest  habitat  types  of  Mon- 
tana. Gen.  Tech.  Rep.  INT-34.  Ogden,  UT:  U.S.  Depart- 
ment of  Agricultiu-e,  Intermountain  Forest  and  Range 
Experiment  Station.  174  p. 

Poelker,  Richard  J.;  Hartwell,  Harry  D.  1973.  Black  bear 
of  Washington.  Biol.  Bull.  14.  Olympia,  WA:  Washing- 
ton State  Game  Department.  180  p. 

Pritchard,  Geoffrey  T.;  Robbins,  Charles  T.  1990.  Diges- 
tive and  metabolic  efficiencies  of  grizzly  and  black 
bears.  Canadian  Journal  of  Zoology.  68:  1645-1651. 

Raine,  R.  M.;  Kansas,  J.  L.  1990.  Black  bear  seasonal  food 
habits  and  distribution  by  elevation  in  Banff"  National 
Park,  Alberta.  In:  International  Conference  on  Bear  Re- 
search and  Management:  Proceedings:  8:  297-304. 

Ronmie,  William  H.;  Turner,  Monica  G.  1991.  Implica- 
tions of  global  climate  change  for  biogeographic  pat- 
terns in  the  greater  Yellowstone  ecosystem.  Conserva- 
tion Biology.  5(3):  373-386. 

Reinhart,  Daniel  P.;  Mattson,  David  J.  1990.  Red  squir- 
rels in  the  whitebark  zone.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-moimtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  256-263. 

Reinhart,  Daniel  P.;  Mattson,  David  J.  1992.  Grizzly  bear 
and  black  bear  habitat  use  in  the  Cooke  City,  Montana, 
area  1990-1991.  Bozeman,  MT:  U.S.  Department  of  the 
Interior,  National  Park  Service,  Interagency  Grizzly 
Bear  Study  Team.  30  p. 

Schallenberger,  A.;  Jonkel,  C.  J.  1980.  Rocky  Mountain 
East  Front  grizzly  studies,  1979.  Annual  Report.  Border 
Grizzly  Project  Special  Report  39.  Missoula,  MT:  Border 
Grizzly  Project.  207  p. 

Schemmel,  R.  1976.  Physiological  considerations  of  lipid 
storage  and  utilization.  American  Zoologist.  16: 
661-670. 

Servheen,  Christopher.  1983.  Grizzly  bear  food  habits, 
movements,  and  habitat  selection  in  the  Mission  Moim- 
tains,  Montana.  Journal  of  Wildlife  Management.  47(4): 
1026-1035. 

Servheen,  Christopher;  Knight,  Richard;  Mattson,  David; 
Mealey,  Steven;  Strickland,  Dale;  Varley,  John;  Weaver, 
John.  1986.  Report  to  the  IGBC  on  the  availability  of 
foods  for  grizzly  bears  in  the  Yellowstone  ecosystem. 
Missoula,  MT:  U.S.  Fish  and  Wildlife  Service,  Office  of 
the  Grizzly  Bear  Recovery  Coordinator.  21  p. 

Steele,  Robert;  Cooper,  Stephen  V.;  Ondov,  David  M.; 
Roberts,  David  W.;  Pfister,  Robert  D.  1983.  Forest 
habitat  types  of  eastern  Idaho-western  Wyoming.  Gen. 
Tech.  Rep.  INT-144.  Ogden,  UT:  U.S.  Department  of 
Agrictdture,  Forest  Service,  Intermountain  Forest  and 
Range  Experiment  Station.  122  p. 

Stirling,  Ian;  Derocher,  Andrew  E.  1990.  Factors  affecting 
the  evolution  and  behavioral  ecology  of  the  modern 


219 


bears.  In:  International  Conference  on  Bear  Research 
and  Management:  Proceedings:  8:  189-204. 

Taylor,  R.  A.  1964.  Coliimbian  ground  squirrel  and  cam- 
bium found  in  grizzly  bear  stomachs  taken  in  the  fall. 
Journal  of  Mammalogy.  45(3):  476-477. 

Tisch,  Edward  L.  1961.  Seasonal  food  habits  of  the  black 
bear  in  the  Whitefish  Range  of  northwest  Montana. 
Missoula,  MT:  University  of  Montana.  108  p.  Thesis. 

Weaver,  T.;  Forcella,  F.  1986.  Cone  production  in  Pmi/s 
albicaulis  forests.  In:  Shearer,  Raymond  C,  comp. 
Proceedings — conifer  tree  seed  in  the  Inland  Mountain 
West  symposium.;  1985  August  5-6;  Missoula,  MT.  Gen. 


Tech.  Rep.  INT-203.  Ogden,  UT:  U.S.  Department  of 
Agricidture,  Forest  Service,  Intermountain  Resesirch 
Station:  68-76. 
Weaver,  T.;  Forcella,  F.;  Dale,  D.  1990.  Stand  develop- 
ment in  whitebark  pine  woodlands.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposiimi  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-moimtain  resom-ce;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station:  151-155. 


220 


ECOLOGICAL  RELATIONSHIP 
BETWEEN  CLARK'S  NUTCRACKER 
AND  FOUR  WINGLESS-SEED  STROBUS 
PINES  OF  WESTERN  NORTH  AMERICA 

Diana  F.  Tomback 


Abstract — The  Clark's  nutcracker  {Nucifraga  columhiana)  is  an 
important  seed  disperser  for  at  least  four  western  North  American 
Strohus  pines:  whitebark  (Pinus  alhicaulis),  limber  (P.  flexilis), 
Colorado  pinyon  {P.  edulis),  and  singleleaf  pinyon  (P.  mono- 
phylla).  Nutcrackers  share  the  role  of  pinyon  pine  seed  disperser 
with  the  pinyon  jay  (Gymnorhinus  cyanocephalus)  and  may  also 
contribute  to  seed  dispersal  of  southwestern  white  pine  (P.  strobi- 
formis).  All  five  species  of  pines  have  large,  wingless  seeds.  In 
whitebark  pine  and  the  pinyon  pines,  ripe  seeds  are  retained  in 
cones. 


In  this  paper  I  briefly  review  the  ecological  relationship 
between  Clark's  nutcracker  {Nucifraga  columhiana)  and 
four  pine  species  of  the  subgenus  Strohus  that  depend  on 
the  nutcracker  for  seed  dispersal.  A  fifth  Strohus  pine, 
southwestern  white  {Pinus  strohiformis)  may  also  be  dis- 
persed by  Clark's  nutcracker  (see  Benkman  and  others 
1984),  but  the  requisite  studies  have  not  yet  been  done. 

This  review  is  not  intended  to  be  exhaustive  but  rather 
to  sketch  the  basics  of  these  interactions,  which  are  simi- 
lar to  those  between  the  the  Eurasian  nutcracker  {N.  caryo- 
catactes)  and  five  European  and  Asian  Strohus  pines. 

GEOGRAPHIC  DISTRffiUTIONS 

The  Clark's  nutcracker  ranges  throughout  much  of 
the  montane  regions  of  the  Western  United  States  (AOU 
1983).  The  eastern  limits  are  the  Black  Hills  of  South 
Dakota  and,  irregularly,  the  pine  forests  of  southwestern 
South  Dakota  and  northwestern  Nebraska,  and  east  of 
the  Rocky  Mountain  firont.  To  the  south,  the  nutcracker 
ranges  into  northern  Baja  California,  and  to  the  north 
into  Canada  along  the  Coast  Mountains  and  Rocky  Moun- 
tains (fig.  1). 

The  geographic  boundaries  of  all  four  Strohus  pines 
known  to  be  dispersed  by  Clark's  nutcrackers  fall  v^dthin 
the  range  of  the  nutcracker  (Critchfield  and  Little  1966) 
(fig.  1).  Whitebark  pine  {Pinus  alhicaulis,  Subsection 
Cembrae)  is  a  subalpine-to-treeline  species  that  occurs 
fi'om  the  Sierra  Nevada  north  through  the  coast  ranges 
and  Northern  Rocky  Mountains  into  Alberta  and  British 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Diana  F.  Tomback  is  Associate  Professor  of  Ecology,  Department  of 
Biology  and  Center  for  Environmental  Sciences,  University  of  Colorado 
at  Denver,  Denver,  CO  80217-3364. 


Columbia.  Limber  pine  {Pinus  flexilis,  Subsection  Strohi), 
which  occupies  a  wide  elevational  range,  is  distributed 
from  southern  California  across  the  Great  Basin  to  the 
Rocky  Mountains  and  north  into  Alberta  and  British 
Columbia.  Both  the  singleleaf  pinyon  {Pinus  monophylla) 
and  Colorado  pinyon  pine  {Pinus  edulis),  which  are  also 
dispersed  by  pinyon  jays  {Gymnorhinus  cyanocephalus) 
(see,  for  example,  Ligon  1978),  are  lower  treeline  species 
in  the  Subsection  Cemhroides.  The  singleleaf  pinyon  pine 
ranges  fi'om  eastern  and  southern  California  south  to 
northern  Baja  California  and  through  the  Great  Basin 
to  western  and  northern  Utah  and  southern  Idaho.  The 
Colorado  pinyon  pine  occurs  in  Utah,  Colorado,  Arizona, 
and  New  Mexico. 


p.  cembroides 


Figure  1 — Geographic  ranges  of  Strobus  pines 
with  large,  wingless  seeds  in  the  Western  United 
States  and  Canada. 


221 


In  addition,  Parry  pinyon  (Pinus  quadrifolia)  and 
Mexican  pinyon  pine  {Pinus  cembroides),  which  are  lower 
elevation  species,  may  occasionally  be  dispersed  by 
Clark's  nutcracker  where  their  ranges  overlap  with  the 
bird,  but  this  has  not  been  documented.  Similarly,  where 
the  range  of  southwestern  white  pine  (Pinus  strobiformis, 
Subsection  Strobi),  overlaps  with  the  nutcracker  in  south- 
ern Colorado,  New  Mexico,  and  Arizona,  its  seeds  may  be 
nutcracker-dispersed  (see,  for  example,  Benkman  and  oth- 
ers 1984);  the  ranges  of  both  Mexican  pinyon  pine  and 
southwestern  white  pine  continue  south  into  Mexico. 

CONE  AND  SEED  TRAITS  OF  PINES 

Large,  wingless  seeds  characterize  the  four  Strobus 
pines  routinely  dispersed  by  Clark's  nutcracker  and 
southwestern  white  pine  (table  1).  In  fact,  the  mean  seed 
weights  of  all  the  wingless-seed  Strobus  pines  that  may 
be  dispersed  by  Clark's  and  Eurasian  nutcrackers  (Nuci- 
fraga  caryocatactes)  (x  =  0.279)  are  significantly  greater 
than  the  mean  seed  weights  of  wind-dispersed  Pinus 
pines  ix  =  0.094)  (Wilcoxon  test,  H  =  15.93,  P  =  0.00006) 
and  also  significantly  greater  than  the  weights  for  the 
winged,  wind-dispersed  Strobus  pines  (x  =  0.057;  H  =  9.31, 
P  =  0.0023,  Tomback  and  Linhart  1990). 

The  cones  of  whitebark  pine  are  indehiscent  and  thus 
do  not  open  after  seeds  have  ripened  (see,  for  example, 
Lanner  1982).  In  the  pinyon  pines,  the  cones  open,  but 
ripe  seeds  are  held  in  depressions  on  cone  scales  by  means 
of  small  flanges  (see,  for  example,  Vander  Wall  and  Balda 
1977).  Both  seed  winglessness  and  seed  retention  in 
cones  increase  the  chances  that  nutcrackers  rather  than 
wind  will  disperse  the  seeds  (Tomback  and  Linhart  1990). 
In  contrast,  the  cones  of  limber  and  southwestern  white 
pine  open  when  the  seeds  are  ripe;  however,  some  limber 
pine  seeds  may  remain  lodged  in  open  cones  as  long  as 
several  weeks  after  ripening  (Tomback  1988;  Torick 
1993).  Discussions  concerning  the  evolution  of  cone  and 
seed  traits  of  bird-dispersed  pines  may  be  found  in  Lanner 
(1980),  Tomback  (1983),  and  Tomback  and  Linhart  (1990). 

The  cones  of  whitebark  pine  (Tomback  1978),  pinyon 
pine  (Vander  Wall  and  Balda  1977),  and  limber  pine 
(Benkman  and  others  1984;  Tomback  and  Kramer  1980) 
ripen  asynchronously  within  and  among  trees  in  late  smn- 
mer,  providing  a  continuous  seed  supply  for  several  weeks 
or  more  and  allowing  nutcrackers  to  disperse  more  of  the 
seed  crop.  In  contrast,  the  cones  of  southwestern  white 
pine  ripen  synchronously  both  within  and  among  trees, 


Table  1 — Mean  seed  weights  of  western  Strobus  pines 
dispersed  by  Clark's  nutcracker.  Data  from 
table  3  in  Tomback  and  Linhart  1 990 


Pine  species  Mean  weight 


Grams 

Singleleaf  pinyon  (Pinus  monophylla)  0.4086 
Colorado  pinyon  (P.  edulis)  .2387 
Whitebark  (P.  albicaulis)  .  1 745 

Southwestern  white  (P.  strobiformis)  .1 680 

Limber  (P.  flexilis)  .0926 


Table  2 — Maximum  sublingual  pouch  capacities  of  Clark's  nut- 
cracker for  four  Strobus  pines.  Scientific  names  of  pines 
as  in  table  1 


Maximum  number 


Pine  species 

of  seeds 

Reference 

Whitebark 

100-150 

Tomback  1982;  Tomback 

and  Knowles  1 989 

Limber 

120 

Vander  Wall  1988 

Colorado  pinyon 

95 

Vander  Wall  and  Balda 

1977 

Singleleaf  pinyon 

38 

Vander  Wall  1988 

which  may  result  in  the  dispersal  of  a  smaller  percentage 
of  the  seed  crop  (Benkman  and  others  1984). 

SEED  HARVESTING  AND  CACHING 
BY  NUTCRACKERS 

The  morphology  and  behavior  of  the  Clark's  nutcracker 
are  similar  to  those  of  the  Eurasian  nutcracker.  The 
Clark's  nutcracker  has  a  long,  sharp  bill  that  is  used  to 
stab  into  and  loosen  closed  cone  scales  or  to  remove  seeds 
from  open  cones  (see  Tomback  1978  for  details  of  seed 
harvesting  behavior).  Nutcrackers  transport  pine  seeds 
in  the  sublingual  pouch,  a  sac  that  is  formed  by  the  floor 
of  the  mouth  and  opens  under  the  tongue  (Bock  and  oth- 
ers 1973).  Maximum  seed  capacities  of  the  pouch  vary 
with  the  pine  species  (table  2).  In  addition,  nutcrackers 
have  a  remarkable  spatial  memory  that  enables  them 
to  relocate  their  seed  stores  accurately  for  more  than 
9  months  after  caching  them  (Balda  and  Kamil  1992; 
Kamil  and  Balda  1985;  Tomback  1980;  Vander  Wall 
1982). 

Clark's  nutcrackers  may  begin  eating  imripe  seeds 
from  the  cones  of  whitebark  (Hutchins  and  Lanner  1982; 
Tomback  1978)  and  limber  pine  (Tomback  and  Taylor 
1987)  as  early  as  mid  to  late  July  in  some  years.  A  bird 
exposes  the  unripe  seeds  by  stabbing  its  bill  between  cone 
scales  and  tearing  off  scale  pieces.  Usually  only  bits  of 
unripe  seeds  can  be  extracted  at  this  time.  Nutcracker- 
damaged  cones  increase  in  frequency  throughout  summer. 
In  limber  pine  they  are  particularly  noticeable:  where 
nutcrackers  removed  seeds,  red-brown  patches  of  frayed 
cone  material  contrast  against  the  green,  undamaged 
cone  scales. 

When  seeds  are  ripe  and  extracted  from  cones  at  a  rela- 
tively rapid  rate  (Hutchins  and  Lanner  1982;  Tomback 
1978;  Vander  Wall  1988),  nutcrackers  begin  to  cache 
them.  Nutcrackers  have  stored  whitebark  pine  seeds 
as  early  as  August  15  in  the  Rocky  Mountains  (Hutchins 
and  Lanner  1982)  and  August  25  in  the  Sierra  Nevada 
(Tomback  1978).  For  limber  pine  and  both  pinyon  pines, 
dates  of  first  seed  storage  are  similar  in  different  montane 
areas:  late  August  to  early  September  (Tomback  and 
Kramer  1980;  Tomback  and  Taylor  1987;  Vander  Wall 
1988;  Vander  Wall  and  Balda  1977). 

Nutcrackers  cache  seeds  at  distances  from  a  few  meters 
to  22  km  from  parent  trees.  Vander  Wall  and  Balda  (1977) 


222 


determined  that  some  nutcrackers  carried  Colorado  pin- 
yon  seeds  7.5  to  22  km  to  caching  areas,  and  Tomback 
(1978)  determined  that  nutcrackers  traveled  as  far  as 
8  to  12  km  with  whitebark  pine  seeds.  More  t3T)ical  dis- 
tances are  shorter  and  up  to  a  few  kilometers  (Hutchins 
and  Lanner  1982;  Tomback  1978;  Vander  Wall  1988). 

In  addition  to  traveling  moderate  to  long  distances, 
nutcrackers  may  fly  to  higher  or  lower  elevations  to  store 
seeds.  For  example,  nutcrackers  in  the  eastern  Sierra 
Nevada  cache  whitebark  pine  seeds  not  only  at  subalpine 
elevations  (2,700  to  3,000  m)  but  also  at  2,100  m  to  2,400  m 
in  the  mixed  Jeffrey  pine  (Pinus  jeffreyi)  and  singleleaf 
pinyon  pine  forest  (Tomback  1978).  Similarly,  nutcrack- 
ers carrying  Colorado  pinyon  seeds  in  northern  Arizona 
may  fly  up  300  m  elevation  to  caching  areas  (Vander  Wall 
and  Balda  1977).  In  the  Colorado  Front  Range,  nutcrack- 
ers were  observed  to  carry  bristlecone  pine  {Pinus  aristata) 
seeds  from  the  subalpine  forest  up  several  hvmdred  meters 
toward  tundra  for  caching  (Baud  1993). 

Clark's  nutcrackers  cache  seeds  both  in  communal 
storage  areas,  often  characterized  by  steep,  south-facing 
slopes  (Tomback  1978;  Vander  Wall  and  Balda  1977),  and 
in  terrain  not  far  from  parent  trees  (Tomback  1978).  The 
Eurasian  nutcracker,  in  contrast,  caches  many  of  its  seeds 
in  a  year-round  territory  (see,  for  example  Mattes  1982). 
Seeds  are  usually  buried  1  to  3  cm  imder  gravelly  soil, 
mineral  soil,  or  duff  (Hutchins  and  Lanner  1982;  Tomback 
1978),  although  a  small  number  may  be  hidden  in  hol- 
lows, under  bark,  in  cracks,  and  in  holes  in  trees  and  in 
logs  (Tomback  1978;  Torick  1993).  Each  cache  typically 
consists  of  one  to  15  seeds,  although  larger  caches  may 
occasionally  be  made;  mean  cache  sizes  are  three  or  four 
seeds  (Tomback  and  Linhart  1990,  table  5  and  references 
therein;  Torick  1993). 

The  specific  kinds  of  sites  selected  by  nutcrackers  for 
biuying  seed  caches  include  the  following:  near  the  base 
of  trees;  around  tree  roots,  fallen  trees,  and  branches; 
around  rocks  and  boulders;  on  rocky,  exposed  ledges; 
within  sparse  to  heavy  plant  cover  and  moss;  at  the  edge 
of  grassy  meadows;  and  within  stands  of  krummholz  pine 
forms  (Hutchins  and  Lannner  1982;  Tomback  1978,  1986; 
Tomback  and  Kramer  1980;  Tomback  and  Taylor  1987; 
Vander  Wall  and  Balda  1977).  In  addition,  nutcrackers 
bury  seeds  at  and  above  treeline  (Baud  1993)  and  in 
burned  forest  terrain  (Tomback  1986;  Tomback  and 
Knowles  1989). 

Estimates  of  how  many  seeds  are  stored  by  a  single 
Clark's  nutcracker  vary  geographically  and  with  the  pine 
species.  For  whitebark  pine,  Tomback  (1982)  estimated 
that  about  35,000  seeds  per  individual  were  cached  after 
a  good  cone  crop  in  the  eastern  Sierra  Nevada,  but  the 
nutcracker  then  cached  the  seeds  of  other  pine  species, 
such  as  Jeffrey  pine  or  singleleaf  pinyon  pine,  or  even  lim- 
ber pine  (Tomback  and  Kramer  1980).  Hutchins  and 
Lanner  (1982)  estimated  that  a  single  nutcracker  cached 
as  many  as  98,000  whitebark  pine  seeds  in  western 
Wyoming.  According  to  Vander  Wall  and  Balda  (1977), 
one  nutcracker  in  northern  Arizona  stored  between  22,000 
and  33,000  Colorado  pinyon  pine  seeds  following  a  good 
cone  crop.  For  northern  Utah,  Vander  Wall  (1988)  calcu- 
lated that  one  nutcracker  stored  a  maximvun  of  16,000 


limber  pine  seeds  and  a  maximimi  of  about  18,000  single- 
leaf  pinyon  seeds. 

In  all  cases  the  stored  seeds  provided  several  times  the 
amoimt  of  energy  required  by  one  nutcracker  during  the 
period  that  the  seed  stores  were  used  (see,  for  example, 
Tomback  1982;  Vander  Wall  and  Balda  1977).  The  excess 
stored  seeds  probably  provided  a  margin  for  loss  of  caches 
to  rodents,  feeding  caches  to  young,  and  some  "forgetting" 
of  seed  cache  location  (see,  for  example,  Balda  and  Kamil 
1992;  Tomback  1982). 

PHENOLOGY  OF  USE  AND  SPECIES 
PREFERENCE 

In  much  of  the  montane  Western  United  States  and 
southern  Canada,  more  than  one  nutcracker-dispersed 
Strobus  pine  occurs  in  the  same  geographic  area  (fig.  1). 
For  example,  in  the  Northern  Rocky  Moimtains  whitebark 
pine  t3T)ically  occurs  at  subalpine  elevations  and  limber 
pine  at  the  lower  forest  boundary,  although  in  some  loca- 
tions the  two  species  may  co-occur  in  the  subalpine  zone. 
In  years  when  both  species  produce  cones,  nutcrackers 
preferentially  harvest  and  store  whitebark  pine  seeds 
before  moving  into  limber  pine  forests  (Tomback  1992). 
Similarly,  in  the  eastern  Sierra  Nevada,  whitebark  pine, 
limber  pine,  and  singleleaf  pinyon  occur  in  the  same  re- 
gion; whitebark  pine  seeds  are  harvested  and  stored 
before  limber  pine  and  singleleaf  pinyon  pine  seeds 
(Tomback  1978;  Tomback  and  Kramer  1980).  In  both 
regions,  nutcrackers  may  prefer  to  take  whitebark  pine 
seeds  first,  because  the  cones  ripen  earlier  than  do  limber 
pine  or  pinyon  pine  cones. 

In  the  Raft  River  Mountains  of  northern  Utah,  stands 
of  limber  pine  and  singleleaf  pinyon  pine  occur  only  a  few 
kilometers  apart.  When  cone  crops  of  both  species  were 
available,  most  nutcrackers  harvested  and  stored  limber 
pine  seeds  before  using  pinyon  pine  seeds  (Vander  Wall 
1988).  Limber  pine  cones  ripened  before  pinyon  pine 
cones  in  this  area.  When  limber  pine  and  Colorado  pin- 
yon pine  both  produced  cone  crops  in  northern  Arizona, 
Vander  Wall  and  Balda  (1977)  observed  nutcrackers  har- 
vesting and  caching  seeds  from  the  closed,  green  cones  of 
both  species  at  the  same  time  in  late  summer.  These  ex- 
amples suggest  that  species  preference  is  probably  based 
on  cone  ripening  sequence. 

Clark's  nutcrackers  are  also  known  to  harvest  and 
store  the  seeds  of  several  wind-dispersed  pine  species  of 
both  the  subgenus  Strobus  and  Pinus.  Stomach  contents 
analysis  of  Clark's  nutcrackers  collected  in  western 
Montana  indicated  that  the  birds  ate,  if  not  cached,  pon- 
derosa  pine  (Pinus  ponderosa)  seeds  and  Douglas-fir 
(Pseudotsuga  menziesii)  seeds  in  addition  to  whitebark 
pine  seeds  (Giimtoli  and  Mewaldt  1978).  Nutcrackers 
regularly  harvest  and  cache  seeds  of  Jeffrey  pine  in  the 
eastern  Sierra  Nevada  (Tomback  1978).  Lanner  (1988) 
noted  nutcracker-damaged  cones  of  Great  Basin  bristle- 
cone  pine  {Pinus  longaeva)  in  the  Wasatch  Range  and 
in  the  Great  Basin.  In  the  Front  Range  of  the  Colorado 
Rockies,  Torick  (1993)  observed  nutcrackers  harvest  and 
cache  ponderosa  pine  seeds,  and  Baud  (1993)  observed 
nutcrackers  harvest  and  transport  bristlecone  pine  seeds 


223 


to  caching  areas.  Baud  (1993)  recently  obtained  experi- 
mental evidence  that  both  bristlecone  pine  and  ponderosa 
pine  seedlings  may  be  routinely  established  from  nut- 
cracker caches.  This  has  been  indirectly  confirmed  by 
Torick  (1993)  who  has  shown  that  the  tree  cluster  growth 
form,  which  is  foimd  in  several  nutcracker-dispersed 
pines  and  arises  from  multiseed  nutcracker  caches 
(Tomback  and  Schuster,  these  proceedings,  £md  references 
therein),  also  occurs  in  ponderosa  and  bristlecone  pine. 

ACKNOWLEDGMENTS 

Support  to  the  author  for  attending  the  SSPE  Work- 
shop was  provided  by  the  College  of  Liberal  Arts  and  Sci- 
ences, University  of  Colorado  at  Denver.  I  thank  Lisa 
Torick  and  Karen  Baud  for  comments  on  an  earlier  draft 
of  this  paper. 

REFERENCES 

American  Ornithologists'  Union.  1983.  Check-list  of  North 
American  birds.  6th  ed.  Lawrence,  KS:  American  Orni- 
thologists' Union.  877  p. 

Balda,  R.  P.;  Kamil,  A.  C.  1992.  Long-term  spatial  memory 
in  Clark's  nutcracker,  Nucifraga  columhiana.  Animal 
Behavior.  44:  761-769. 

Baud,  K.  S.  1993.  Germination  rates  and  losses  to  ro- 
dents of  artificial  nutcracker  caches  of  three  species  of 
Pinus  placed  in  three  elevational  zones  in  the  Colorado 
Front  Range.  Data  on  file.  Department  of  Biology,  Uni- 
versity of  Colorado  at  Denver. 

Benkman,  C.  W.;  Balda,  R.  P.;  Smith,  C.  C.  1984.  Adapta- 
tions for  seed  dispersal  and  the  compromises  due  to 
seed  predation  in  limber  pine.  Ecology.  65:  632-642. 

Bock,  W.  J.;  Balda,  R.  P.;  Vander  Wall,  S.  B.  1973.  Mor- 
phology of  the  sublingual  pouch  and  tongue  muscula- 
ture in  Clark's  nutcracker.  Auk.  90:  491-519. 

Critchfield,  W.  F.;  Little,  E.  L.,  Jr.  1966.  Geographic 
distribution  of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture,  For- 
est Service.  97  p. 

Giuntoli,  M.;  Mewaldt,  L.  R.  1978.  Stomach  contents 
of  Clark's  nutcracker  collected  in  western  Montana. 
Auk.  95:  595-598. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Kamil,  A.  C;  Balda,  R.  P.  1985.  Cache  recovery  and  spa- 
tial memory  in  Clark's  nutcrackers  (Nucifraga  Colum- 
biana). Journal  of  Experimental  Psychology:  Animal 
Behavior  Processes.  11:  95-111. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor 
in  the  ecology  and  evolution  of  limber  £ind  whitebark 
pines.  In:  Dancik,  B.  P.;  Higginbotham,  K.  O.,  eds.  Sixth 
North  American  forest  biology  workshop:  proceedings. 
Edmonton,  AB:  University  of  Alberta:  15-48. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for 
seed  dispersal  by  Clark's  nutcracker.  Canadian  Journal 
of  Forest  Research.  12:  391-402. 


Lanner,  R.  M.  1988.  Dependence  of  Great  Basin  bristle- 
cone pine  on  Clark's  nutcracker  for  regeneration  at  high 
elevations.  Arctic  and  Alpine  Research.  20:  358-362. 

Ligon,  J.  D.  1978.  Reproductive  interdependence  of 
pinon  jays  and  pinon  pines.  Ecological  Monographs. 
48:  111-126. 

Mattes,  H.  1982.  Die  Lebengemeinschaft  von  Tannenhaher 
imd  Arve.  Birmensdorf,  Switzerland:  Swiss  Federal  In- 
stitute of  Forestry  Research.  74  p. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. Living  Bird.  16:  123-161. 

Tomback,  D.  F.  1980.  How  nutcrackers  find  their  seed 
stores.  Condor.  82:  10-19. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.  1983.  Nutcrackers  and  pines:  coevolution 
or  coadaptation?  In:  Nitecki,  M.  H.,  ed.  Coevolution. 
Chicago:  University  of  Chicago  Press:  179-223. 

Tomback,  D.  F.  1986.  Post-fire  regeneration  of  krummholz 
whitebark  pine:  a  consequence  of  nutcracker  seed  cach- 
ing. Madrono.  33:  100-110. 

Tomback,  D.  F.  1988.  Nutcracker-pine  mutualisms:  multi- 
trunk  trees  and  seed  size.  In:  Ouellet,  H.,  ed.  Acta  XIX 
Congressus  Internationalis  Ornithologici,  Vol.  I;  1986 
June  22-29.  Ottawa,  ON:  University  of  Ottawa  Press: 
518-527. 

Tomback,  D.  F.  1992.  Observations  in  the  Wind  River 
Range.  Unpublished  data  on  file,  Department  of  Biol- 
ogy, University  of  Colorado  at  Denver. 

Tomback,  D.  F.;  Knowles,  J.  W.  1989.  Post-fire  whitebark 
pine  seed  dispersal  by  Clark's  nutcracker  in  Yellowstone 
National  Park.  Unpublished  data  on  file,  Department 
of  Biology,  University  of  Colorado  at  Denver. 

Tomback,  D.  F.;  Kramer,  K.  A.  1980.  Limber  pine  seed 
harvest  by  Clark's  nutcracker  in  the  Sierra  Nevada: 
timing  and  foraging  behavior.  Condor.  82:  467-468. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Tomback,  D.  F.;  Taylor,  C.  L.  1987.  Tourist  impact  on 
Clark's  nutcracker  foraging  activities  in  Rocky  Moim- 
tain  National  Park.  In:  Singer,  F.,  ed.  Wildlife  manage- 
ment and  habitats.  Washington,  DC:  National  Park 
Service  and  George  Wright  Society:  158-172. 

Torick,  L.  L.  1993.  Field  and  laboratory  evidence  for 
regeneration  of  wind-dispersed  pines  from  animal- 
dispersed  caches.  Data  on  file,  Department  of  Biology, 
University  of  Colorado  at  Denver. 

Vander  Wall,  S.  B.  1982.  An  experimental  analysis  of 
cache  recovery  in  Clark's  nutcracker.  Animal  Behavior. 
30:  84-94. 

Vander  Wall,  S.  B.  1988.  Foraging  of  Clark's  nutcrackers 
on  rapidly  changing  pine  seed  resources.  Condor.  90: 
621-631. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations 
of  Clark's  nutcracker  and  the  pinon  pine  for  efficient 
seed  harvest  and  dispersal.  Ecological  Monographs. 
47:  89-111. 


224 


Forest  Structure 
and  Dynamics 


International  Workshop 
St.  Moritz  1992 


STRUCTURE  OF  SWISS  STONE  PINE 
STANDS  IN  NORTHEASTERN  ITALY 


Giovanna  De  Mas 
Elena  Piutti 


Abstract — This  contribution  suggests  a  structural  typology  for 
Swiss  stone  pine  {Pinus  cembra)  stands  in  the  eastern  range  of 
the  ItaUan  Alps.  Study  of  the  structure  in  these  stands  makes  it 
possible  to  understand  their  dynamics  and  to  evaluate  environ- 
mental and  scenic  aspects  of  the  present  expansion  of  high- 
elevation  forests.  This  paper  describes  eight  structural  types 
that  show  different  rhythms  in  the  forest  establishment  process. 


The  structure  of  Swiss  stone  pine  {Pinus  cembra  L.) 
stands  has  been  studied  by  several  authors,  although  with 
different  methodological  approaches  (see,  for  examples, 
Contini  and  Lavarelo  1981,  1982;  Del  Favero  and  others 
1985;  Kuoch  1972;  Mayer  and  Ott  1991;  Piussi  and 
Schneider  1985;  Rachoy  1976;  Schiechtl  and  Stern  1975, 
1979,  1983,  1984;  Stern  and  Helm  1979).  These  studies 
are  of  particular  interest.  They  allow  one  to  make  an 
improved  assessment  of  management  and  environmental 
implications  resulting  from  reforestation  of  high-elevation 
abandoned  pastures.  These  have  become  a  typical  feature 
of  many  gdpine  areas  as  a  consequence  of  changed  economic 
structures. 

In  fact,  since  the  structure  analysis  of  a  tree  stand  pro- 
vides a  detailed  description  of  the  spatial  and  temporal 
tree  distribution  pattern,  these  early  approaches  are  nec- 
essary to  understand  the  complex  mechanisms  that  con- 
trol natural  regeneration  and  intra-  and  interspecific  com- 
petition. Moreover,  they  enable  us  to  quantify  rhythms  of 
different  tree  species  establishment  and  to  define  the  com- 
plex stand-stability  degree.  The  objective  of  this  investi- 
gation, which  is  part  of  the  INTEGRAL?  project  finan- 
cially supported  by  EEC,  is  to  detect  early  structure  types 
of  stone  pine  stands  on  the  southern  slope  of  the  eastern 
Alps.  Knowledge  of  these  stands  is  comparatively  scanty. 

FIELD  SURVEYS 

Field  work  was  carried  out  in  a  large  part  of  the  autono- 
mous Province  of  Bolzano  and  in  the  Province  of  Belluno. 
In  particular,  the  study  was  conducted  in  stone  pine 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Giovanna  De  Mas  is  a  Researcher  and  Elena  Piutti  is  making  her  re- 
search doctorate  stage.  Land  and  Agro-Forest  Environment  Department, 
University  of  Padova,  35100  Padova,  Italy. 


coenoses  in  the  following  valleys  of  Bolzano  Province: 
Venosta,  Passiria,  Sarentino,  Funes,  Luson,  Gardena, 
and  Badia  and  in  Cortina  basin  in  Belluno  Province.  This 
scenic  area  exhibits  a  remarkable  climatic  and  geologic 
variability. 

Climate  becomes  increasingly  continental  while  mov- 
ing fi*om  east  to  west  (Del  Favero  and  others  1985;  Fliri 
1975).  Temperatures  are  moderate  in  the  east  and  an- 
nual precipitation  exceeds  1,000  mm,  with  rainfall  max- 
ima in  spring  and  autumn.  In  the  west,  climate  gradually 
becomes  colder  in  winter,  and  summers  are  dry  and  hot. 
Annual  precipitation  is  less  than  1,000  mm  and  the  rain- 
fall regime  is  continental.  Geological  substrata  cheinge 
fi'om  limestone  in  the  east  to  silicate  bedrock  in  the  west. 

In  the  eastern  part,  stone  pine  is  found  at  the  eastern 
periphery  of  its  natural  range.  It  forms  pure  or  mixed  for- 
ests associated  with  larch  (Larix  decidua  Mill.)  and/or 
spruce  (Picea  abies  Karst.).  It  is  occasionally  mixed  with 
Scots  pine  (Pinus  sylvestris  L.)  or  with  silver  fir  (Abies 
alba  Mill.).  The  vegetational  aspects  of  these  associations 
were  described  by  Filipello  and  others  (1976;  1980-81). 

Based  on  the  results  of  field  surveys  and  on  data  avail- 
able from  forest  management  plans,  it  was  possible  to 
make  a  preliminary  early  subdivision  of  the  stone  pine 
stands  examined  into  comparatively  homogeneous  struc- 
ture categories.  These  categories  range  fi*om  open  stands 
of  remarkably  large-size  trees,  sometimes  accompanied  by 
recent  young  trees,  to  more  or  less  closed  forests. 

First,  a  rough  classification  was  conceived.  Then,  some 
forest  stands  were  subjectively  identified  by  size  and  vari- 
ability of  the  category  itself  These  stands  were  consid- 
ered to  approximately  represent  the  specific  structure 
conditions.  In  each  of  these  stands,  a  rectangular  400-m'^ 
sample  plot  was  marked.  In  this  plot,  the  following  pa- 
rameters were  recorded  for  each  tree  exceeding  a  diam- 
eter of  7.5  cm  at  ground  level:  breast  height  diameter, 
height,  age  at  tree  base,  distance  of  tree  center  fi"om  the 
two  plot  vertices.  Then,  spatial  tree  distribution  (vertical 
and  horizontal)  was  drawn  with  early  processing  and  the 
following  stand  parameters  were  computed  per  hectare: 
niunber  of  trees,  basal  area,  voliune  (determined  using 
the  double  entry  volvune  tables  of  the  Nationsd  Forest 
Inventory,  I.S.A.F.A.,  1984),  average  diameter,  and  aver- 
age height. 

In  total,  39  sample  plots  were  surveyed.  Their  main 
characteristics  are  summarized  in  table  1.  Subjective 
sampling  and  the  selection  of  a  comparatively  low  number 
of  small  sample  plots  was  inevitable  because  of  limited 
funding  and  in  view  of  the  objective  of  analyzing  at  least 
the  most  important  stands  among  those  observed. 


226 


°  12  ffl 
o  o  S 

<  o 


It 

'x  o 


>  « 
<  =6 


0) 

E 

3 
O 
> 


C 

o 
o 

Q. 

E 
o 
o 


u 

3 
i_ 

a 

CO 


o  .£ 


_  « 

(0  £ 

(0  m 

»  i 

CD 

O  € 

.  V) 


o 
o 

Q. 
(A 
< 


C 

.2 

(0 

> 
o 


Q-  3 

c 


52 

CO 

:2 


■2 

I 


■2 

I 


a5TtcoLn-i--^oot^(r)OLntnino^-cooinooooir)OP5t^or3cr)LOLr>LnoooooT-rs. 
r~~.cO'r-o5Qot^r^-<tcor^cvjcor>»t>~.ocoooino)oot^-t-coi^O'«*co<oo<35'a-oin-r-c\j-i-i^coo 

■"-■^CM        •t-i--.-OJCVJ-.-CVJC\Ji-i-i-i-T-T-T-CM't-CJi-'>-^C\JC0t--t-  t-C\jT-C\JCVJCVJC\Ji-CVJ 


o 

in 

IT) 

Lf) 

IT) 

O 

o 

m 

o 

in 

o 

o 

o 

ir> 

o 

o 

q 

q 

CM 

00 

00 

00 

o 

o 

in 

in 

o 

in 

o 

o 

O 

in 

o 

in 

o 

q 

CD 

00 

CO 

CO 

CM 

CO 

CO 

CO 

CO 

CO 

CO 

cri 

CM 

CO 

CO 

CO 

in 

CD 

CO 

IT) 

iri 

CD 

CO 

CD 

CM 

CM 

C\J 

CM 

CM 

CVi 

CO 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

OO-i-CMLnCMQOh-O 

TtcjjcdcDCMooT-^CMin 


T-Lncoqqqi--r..^in^ 
o'cicoinTtodi^ihT^csco 


CO  CO      C33  in  o 

CO         CM  CD  CM  CM  T-^ 


cDi-coor^coinc7)ooincDCM 

T-'cDT-^obodcOCMOOCMCjicjiod 


CO 

00 

CO  O) 

CO 

<J> 

00 

00 

o 

in 

in 

CM 

m 

G) 

CO 

00 

00 

CM 

o 

CO 

Oi 

CO 

o 

<J> 

CM 

CM 

in 

CM 

CM 
CM 

d 

CO 

CO 
CM 

CO 
CM 

CO 
CM 

00 
CM 

in 

CM 

ai 

CM 

ai 

CM 

ai 

d 

CM 

1^ 

d 

CM 

CM 

d 

CM 

in 

CM 
CM 

00 
CM 

d 

CO 

CO 
CM 

CM 

5 

CM 

CM 

in 

oci 

CM 

CD 
CM 

CO 

CO 

in 

CO 

00 
CM 

CO 

co 

CO 

in 

q 

T— 

00 

CD 

o 

in 

oo 

C3> 

q 

CO 

cn 

CM 

q 

1^ 

00 

CD 

N. 

o 

CM 

co 

CJ) 

in 

CM 

in 

q 

CM 

CD 

CO 

00 

d 

00 

CD 

CO 

d 

d 

00 

CD 

d 

00 

Cli 

00 

N.' 

CD 

d 

iri 

00 

in 

CO 

CD 

■>*■ 

CD 

in 

in 

00 

CO 

CO 

CM 

CO 

Cvj 

CM 

CM 

in 

CJ) 

in 

CO 

in 

00 

CO 

00 

CM 

CM 

00 

00 

in 

CM 

CO 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CO 

CO 

CO 

in 

CM 

1- 

CO  1- 


CJ)  o 
in  in  CO 


00 


in  CM 

CO  CM 


CD 
CM 


in  <3)  in 
^  m  r-. 


CM 


<M 


CO  O) 
CD  CO 


CM 

in 


m  o  CO 


CO 


00 
CO 


O  00  CM 

in  t-  CO 


o 
in 


CO  T-  CO 
CM 


CM  CO  CO  in 


Ti-  CM  r--  CO 


CO  CM  O 
CO  CM 


o 

CO 

O 

O) 

00 

O 

IT) 

O 

O) 

O 

CO 

o 

<J) 

CM 

1^ 

00 

in 

a> 

CO 

CM 

(0 

00 

CO 

00 

CO 

00 

o 

CM 

l>. 

O 

CO 

CM 

CM 

C7> 

in 

CO 

CM 

O 

c:^ 

CT) 

1^ 

00 

CO 

00 

CO 

00 

(0 

0^ 

00 

CO 

CM 

CD 

CO 

CO 

CD 

CM 

■t 

CD 

o 

00 

CM 

Oi 

O) 

00 

00 

00 

CM 

o 

o 

in 

q 

o 

CO 

q 

CO 

in 

CO 

d 

CD 

CM 

CM 

00 

d 

CO 

CO 

C> 

00 

CM 

00 

00 

CO 

0C3 

d 

in 

CO 

iri 

d 

CO 

CD 

CM 

•<t 

in 

in 

in 

CM 

CO 

CM 

CM 

CO 

CM 

CO 

CO 

CO 

CM 

CO 

CM 

CM 

CM 

CM 

CO 

CM 

CO 

CO 

in 

CM 

CM 

ir> 

in 

CM 

CO 

oinininoinininininin 
inr^cMr^int^^cMCMi^cMCM 
inincDincMin'*'<tin't'«t 


oinooinooinoooinininin 
inr^inincMininN-^oor^cMCMCM 
■^•<tcDLncoor^oO'-cocooO'<!tcocM 


o  in 

O  CM 


oininininininoinoo 
or^cMr^cMi^r-mr^oo 
cDinincDtcom'tcO'^'a- 


LU 

C/)  C/D  C/D  LU 


Z  Z  Z  Z  Z  lU  z 


HI 

O  LU  LU  O  z 
CO  CO  CO  CO  LU  LU 


LU  ^  W  ^  Uj  ^  ^ 
ZCO^ZZZCOZ 


^  ^  ^ 

LULULULUmZZZ 
ZZZZZZZZ 


^  LU  LU 

^  z  z 

Z  Z  Z  LU  UJ 


o 

o 

o 

o 

O 

o 

o 

o 

O 

o 

o 

in 

o 

o 

O 

o 

O 

o 

o 

O 

O 

O 

o 

o 

o 

o 

O 

o 

in 

o 

in 

o 

o 

o 

o 

o 

o 

o 

in 

in 

o 

o 

O 

in 

CM 

in 

00 

in 

CO 

CM 

in 

o 

(0 

in 

o 

CD 

CD 

■t 

in 

o 

in 

in 

o 

in 

in 

CM 

o 

in 

in 

00 

00 

00 

in 

in 

CJ) 

CO 

O) 

O 

00 

o 

T— 

CD 

o 

CM 

oo 

CM 

00 

o 

o 

o 

o 

o 

CO 

00 

00 

in 

CO 

<3) 

00 

CO 

00 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CM 

CO{OCOCO(0(OOBCOD3 

CDCDOOCDCDCQCDU.U. 


3   3    3  3 


SOSOOOOOSOOOSCOCOCOCOCOCOCOCOCOOO 
-l>>>>>>>>C0C0C0C0Q.CLC0C0C0C0C0C0C0OO 


T-cMco'^incor~-.ooc3>oi-CMco'<i-incDr^ 


cooiOi-CMco^incoi^oocj)Oi-cMco^incor^ooo> 

i-i-CMCMCMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCOCO 


227 


METHODS 

As  mentioned  earlier,  the  structural  analysis  of  a  forest 
stand  enables  us  to  describe  its  appearance  and  to  under- 
stand its  dynamics — the  patterns  and  the  mechanisms 
that  have  controlled  its  development  until  now.  The  pres- 
ent state  results  from  the  interaction  between  natiiral 
components  and  human  disturbances. 

The  structure  of  a  forest  coenosis  can  be  illustrated  by 
means  of  its  characteristics  or  by  "chronological"  ele- 
ments. The  former  can  be  sampled  and  observed  more 
easily.  They  allow  a  static  view  of  stand  conditions  that 
can  be  described  in  detail.  For  this  purpose,  the  following 
elements  are  considered  in  general:  composition  (numeric 
and  volumetric),  percentage  of  tree  nimiber  in  diameter 
classes  (usually  5-cm  range)  and  in  height  classes  (of 
different  width),  related  to  total  sum  of  trees,  numeric 
indices  of  spatial  distribution,  and  some  dendrometric 
parameters. 

In  this  investigation,  the  elements  given  were  taken 
into  consideration  for  each  sample  plot.  However,  only 
the  following  ones  were  used  in  the  assessment  of  struc- 
tural types:  numeric  composition,  diameter  classes,  rela- 
tive height  classes  related  to  three  layers,  and  ratio  of 
variance  with  mean  of  niunber  of  trees  found  in  the  four 
subplots,  each  100  m^  wide,  into  which  the  sample  plot 
had  been  divided. 

The  chronological  structure  of  a  stand  provides  a  pic- 
ture of  the  stand  dynamics  and  of  species  establishment. 
Thus  the  chronological  aspect  is  essential  for  structure 
analysis,  although  it  is  often  neglected  because  of  survey- 
ing difficulties.  Its  identification  is  made  by  subdividing 
the  number  of  trees,  distinguished  by  tree  species,  into 
chronological  classes  with  an  unstandardized  range  (in 
the  case  considered  in  this  work,  a  20  years'  lapse  rate 
was  fixed). 

To  develop  an  early  structural  typology  for  the  stands 
investigated,  we  tried  to  aggregate  sample  plots  into  ho- 
mogeneous groups.  Distance  measures  were  computed 
using  Minkowski's  algorithm  (Franceschetti  and  Provasi 
1978)  taking  the  physical  and  chronological  factors  into 
consideration,  after  having  standardized  data  and  con- 
verting them  into  percentages  of  the  totals. 

Because  of  the  unsatisfactory  results  obtained  in  this 
early  processing,  another  method  was  tried.  Three  sets  of 
sample  plots,  different  in  composition,  were  selected.  The 
first  set  (17  sample  plots)  includes  those  where  stone  pine 
represents  over  80  percent  of  the  species  composition;  the 
second  (14  plots)  includes  those  where  stone  pine  is 
present  by  less  than  80  percent;  spruce  prevails,  or  is  the 
second  species  by  importance.  The  third  set  (eight  plots) 
comprises  those  where  larch  prevails  or  is  the  second  spe- 
cies after  stone  pine  (which  is  in  any  case  present  by  less 
than  80  percent).  Then,  the  distance  measure  computa- 
tion was  repeated  in  each  of  these  three  sets,  considering 
as  factors  only  the  ones  concerning  the  physical  structure. 

Before  discussing  the  results,  it  is  necessary  to  state 
that  no  case-by-case  assessment  could  be  made  of  the 
modifications  that  cutting  or  grazing  caused  on  the  struc- 
ture of  the  single  sample  plots,  because  they  were  hardly 


noticeable  on  small-size  sireas.  It  is  only  possible  to  be- 
lieve that  both  activities  were  carried  on  in  the  areas  in- 
vestigated and  the  dynamism  that  led  to  the  present 
structure  was  affected  by  severe  cuts  made  in  the  past 
and  by  the  progressive  abandonment  of  grazing  practices. 

RESULTS 

Five  stand  structure  types  (fig.  1)  are  distinguished  in 
the  first  set  of  sample  areas  where  stone  pine  is  present 
by  over  80  percent.  Three  types,  however,  are  based  on 
one  observation  only. 

The  first  type  (fig.  2,  structural  type  1.1),  includes  six 
sample  plots.  Diameter  distribution  decreases  as  diam- 
eter increases,  with  trees  that  reach  the  55-cm  class.  As 
to  vertical  distribution,  a  remarkable  concentration  of 
heights  is  foimd  in  the  intermediate  layer,  although  a  fair 
share  occurs  in  the  upper  third. 

In  the  second  type  (fig.  3,  structural  type  1.2,  eight 
sample  plots),  the  regeneration  process  occurred  gradu- 
ally as  tree  age  ranges  from  20  to  220  years.  Most  of 
them  are  101  to  120  years  old,  while  a  few  exceed  140 
years.  Diameter  distribution  is  different;  it  is  unimodal 
with  a  maximum  in  the  20-cm  class  and  with  a  high  con- 
centration in  the  15-  to  30-cm  classes.  Vertical  distribu- 
tion shows  most  heights  in  the  upper  third. 

In  the  third  type  (fig.  4,  structural  type  1.3),  the  stand 
has  a  reduced  niunber  of  trees  per  hectare  (140  trees/ha), 
and  trees  are  missing  at  the  youngest  age.  The  last  new 
generation  occurred  60  to  140  years  ago.  The  height  dis- 
tribution of  trees  appears  well  balanced  in  the  layers, 
even  if  a  slight  trend  to  one-storied  formation  can  be  no- 
ticed. 

Contrary  to  that,  in  the  fourth  type  (fig.  5,  structural 
type  1.4),  regeneration  appears  to  be  more  recent  because 
the  new  generation  comprises  trees  40  to  80  years  old. 
Tree  heights  concentrate  in  the  lower  third  just  as  do  the 
diameters,  which  are  mainly  foimd  within  the  10-  to  20- 
cm  classes. 

The  last  type  (fig.  6,  structural  type  1.5)  represents  the 
structural  conditions  of  a  comparatively  rare  pure  even- 
aged  stone  pine  stand  where  all  mature  individuals  are 
over  220  years  old  while  younger  trees  are  missing.  Trees 
(400/ha)  have  diameters  varying  from  30  to  60  cm  and 
heights  concentrated  in  the  third  upper  layer. 

Three  types  are  distinguished  in  the  second  set.  They 
concern  sample  plots  where  stone  pine  is  less  than  80  per- 
cent and  spruce  is  either  prevailing  or  the  second  species. 

In  the  first  type  (fig.  7,  structural  type  2.1),  including 
five  sample  plots,  diameters  range  fi-om  10  to  50  cm  and 
exhibit  a  bimodal-type  pattern  with  a  main  maximum  in 
the  30-cm  class  and  a  secondary  maximum  in  the  20-cm 
class.  Trees  are  concentrated  in  the  20-  to  35-cm  classes. 
The  high  number  of  trees  in  the  10-cm  class  (youngest 
individuals)  should  be  emphasized;  this  reflects  well- 
established  regeneration.  As  to  vertical  distribution,  a 
trend  to  one-storied  stand  formation  is  evident  since  over 
60  percent  of  the  trees  are  included  in  the  upper  third. 

From  the  chronological  point  of  view,  a  slight  imeven- 
age  structure  is  apparent.  It  is  characterized  by  a  wide 
range  of  tree  ages  varjdng  from  20  to  over  220  years. 


228 


200  t 


150 


100 


50  ■ 


18  30  16  15  29  19  4   23  21  32  10  12  26  33  20  25  27 

Plot  Numbers  (table  1) 


200  T 


150  - 


100  - 


50 


B 


37  35  13  3    9  14  1    2  31  17  34  36  39  5 

Plot  Numbers  (table  1) 


28  8    6  38  7  22  11  24 


Plot  Numbers  (table  1) 


Figure  1 — Dendrograms  obtained  from  the  measurements  of  distance  among  the  elements  of 
physical  structure.  A,  set  of  stone  pine  plots;  B,  set  of  stone  pine  and  spruce  plots;  C,  set  of 
stone  pine  and  larch  plots. 


229 


30- 


25- 


e 

V 


■a 
13 


21-40      41-60      61-HO     81-IOU    101-120    I2I-I4U    MMM)    lAMW    181  21X1   201-220  >22\ 


chronological  age  classes 


I     I  P.  cembra 


Figure  2— Structural  type  1.1;  including  plots  1 8, 
30, 16, 15,  29,  and  19.  The  main  characteristics  of 
the  structural  type  are  illustrated  by  average  values 
of  physical  and  chronological  elements.  In  particu- 
lar, the  upper  part  of  the  figure  shows  tree  fre- 
quency in  diameter  classes  and  the  value  of  the 
average  diameter.  A  second  graph  shows  percent- 
age of  trees  related  to  three  layers.  These  are  sub- 
divided by  maximum  height  and  represent  the 
vertical  type  structure.  In  addition,  average  height 
and  maximum  height  are  shown.  A  third  graph  dis- 
plays the  percentage  of  trees,  distinguished  by  tree 
species,  into  age  classes  of  20  years'  width.  Fi- 
nally, a  list  is  given  of  some  average  dendrometric 
parameters  that  are  related  to  the  hectare  (number 
of  subjects,  basal  area,  volume)  and  the  number  of 
plots  where  the  spatial  distribution  index  is  less 
than  or  more  than  one,  respectively.  For  this  last 
parameter,  which  in  any  case  never  affected  type 
formation  in  a  significant  way,  the  average  value 
was  not  computed  because  it  often  proved  to  be 
highly  variable  inside  the  group. 


230 


average  h  =  13  m     max.  h  =  17.6  m 


relative  frequency 


30- 


 T  T  r  T  Y  Y  y  Y  T  Y- 

21-40      41-«      61-80     Hl-lOn    1(11-120    12!-l<0    141-l«)    161  IHO    181-2(0  201-221)  >221 


chronological  age  classes 


I     I  P.  cembra 


Figure  3— Structural  type  1 .2;  including  plots  4,  23,  21 , 
32, 10, 12,26,  and  33. 


231 


CO 

d 
e 


10  15  20  25  30  35  40  45  50  55  60  65  70  75 
diameter  classes  (cm) 


60- 


chronological  age  classes 


P.  cembra 


Figure  4— Structural  type  1 .3;  including  plot  20. 


232 


400 


10  15  20  25  30  35  40  45  50  55  60  65  70  75 
diameter  classes  (cm) 


average  h  =  12.9  m     max.  h  =  23  m 


0       10      20      30      40      50      60      70      80  90 


relative  frequency 


«5- 


 T  T  1  1  r—  1  1  1  

21-40      41-60      (ilSi      81-1(10    101-121)    12M40    141-160    lAI-lM}    lill-2in   201-Z20  >221 


chronological  age  classes 


I     I  P.  cembra 


Figure  5— Structural  type  1 .4;  including  plot  25. 


233 


10  15  20  25  30  35  40  45  50  55  60  65  70  75 
diameter  classes  (cm) 


average  h  =  21  m     max.  h  =  23.5  m 


15.8-23.5 


M  7.9-15.7 


J3 


0-7.8 


relative  frequency 


0       10      20      30^     40      50      60      70      80  90 


-1  1  1  1  1  1  r 

21-4(1      4I'M)      61-80     81100    101  120    12I-I«I    141-160   161180   IRl-200  2U1-220  >221 


No/ha  400 
G/ha  m  55 
V/ha  m  494.1 
f .  spatial  index 
<1  1 
>1  0 


chronological  age  classes 


I     I  P.  cembra 


Figure  6— Structural  type  1 .5;  including  plot  27. 


234 


400 


350- 
300 
^  250- 
i  200- 


average  dbh  =  29.4  cm 


150 
100 
50 
0 


10  15  20  25  30  35  40  45  50  55  60  65  70  75 
diameter  classes  (cm) 


average  h  =  17.3  m     max.h  =  22.1 


0       10      20      30      40      50      60      70      80  90 


relative  frequency 


 T  T  f  1  1  T  T- 

21-40      41.00      A|.aO     81-1(111    101.120    121  140    I4MM    lAI-IM)    181.2(10   201.221)  >22l 


chronological  age  classes 


P.  cembra 


L.  decidua 


P.  abies 


Figure  7— Structural  type  2.1 ;  including  plots  37,  35, 
13,  3,  and  9. 


I 


235 


There  is,  however,  a  large  quantity  of  trees  in  the  121-  to 
160-year  classes,  with  a  maximum  in  the  class  of  121  to 
140  years.  Moreover,  larch  is  conspicuous  only  in  the  ad- 
vanced age  classes  (from  120  years  and  older,  with  a 
maximum  in  the  201-  to  220-year  class).  In  this  type,  the 
average  nimaber  of  trees  per  hectare  amounts  to  625, 
basal  area  per  hectare  is  42.5  m^,  while  average  volimae 
per  hectare  turns  out  to  be  347.2  m^. 

The  second  type  (fig.  8,  structural  type  2.2),  including 
four  sample  plots,  shows  a  wider  diameter  range  (from 
10  to  70  cm).  A  trend  can  be  clearly  noticed  for  unimodal 
distribution  with  a  maximum  in  the  40-cm  class.  A  fair 
variability  can  be  observed  also  in  the  upright  tree  distri- 
bution. In  fact,  individuals  with  heights  in  the  upper 
layer  prevail;  nevertheless  individuals  in  the  lower  third 
are  well  represented. 

A  slight  uneven-age  condition  can  also  be  noticed  in  this 
second  type  (ages  range  from  20  to  220  years),  but  age  dis- 
tribution is  more  regular  with  141-  to  160-year  classes 
prevailing.  The  average  values  are  369  trees  per  hectare, 
basal  area  is  35.2  m^,  and  volume  is  329.2  m^. 

The  last  type  of  the  second  group  (fig.  9,  structural  type 
2.3)  includes  five  sample  plots  and,  in  contrast  with  those 
so  far  examined,  it  has  a  decreasing  diameter  distribution 
with  a  high  nimaber  of  small-  or  middle-size  trees  that  are 
not  strictly  found  in  the  yoimger  chronological  classes. 
Height  distribution  is  more  uniform.  There  is  a  larger 
one-storied  formation,  because  nearly  90  percent  of  the 
trees  form  the  two  upper  layers.  Moreover,  differences  in 
age  are  slightly  less  conspicuous  (40  to  200  years).  Most 
trees  are  in  the  80  to  140  classes.  The  number  of  trees 
per  hectare  (670)  is  comparatively  high,  while  the  many 
small-size  trees  lower  the  value  for  basal  area  (21.8  m^) 
and  for  volrnne  (172.8  m^)  per  hectare.  The  decreased  pro- 
portion of  larch  is  emphasized  and  confirms  what  already 
has  been  stated.  Larch  less  than  80  years  is  not  repre- 
sented in  this  case;  instead  other  species  (Scots  pine  and 
silver  fir)  join  in  the  formation. 

In  the  third  set,  including  areas  where  stone  pine  is 
represented  by  less  than  80  percent  and  larch  is  either 
predominant  or  the  second  species,  it  should  be  noted  that 
aggregations  obtained  through  distance  measures  do  not 
seem  to  be  satisfactory  because  of  extreme  stand  differ- 
ences. Consequently,  it  was  impossible  to  single  out 
groups  representing  average  conditions.  It  may  be  that 
only  a  subsequent  closer  investigation  with  an  increased 
number  of  observations  will  enable  us  to  define  a  struc- 
tural typology  for  these  coenoses.  Larch  is  found  either  in 
the  oldest  chronological  classes  or  in  the  young  classes, 
where  structures  are  more  open. 


years  can  be  observed.  Most  of  the  stsind,  however,  con- 
sists of  trees  with  ages  ranging  fi-om  20  to  120  years,  but 
with  a  peak  in  the  60-  to  80-year  class.  Thus  it  can  be 
supposed  that  in  this  case  regeneration  has  occurred 
within  a  comparatively  short  time,  and  is  now  decreasing 
gradually.  The  stands  have  a  good  number  of  trees  per 
hectare  (662).  Conversely,  the  other  dendrometric  param- 
eters (all  of  them  rather  low)  suggest  that  biospace  will  be 
saturated  chiefly  by  a  growth  increase  of  individuals  al- 
ready present  rather  than  by  additional  trees. 

Conversely,  in  the  second  type  the  physical  and  chrono- 
logical aspects  and  the  dendrometric  parameters,  which 
are  higher  than  in  the  previous  type  (531  trees/ha, 
25.6  m^/ha  basal  area,  179.1  m^  mass-volimie-stock/ha) 
suggest  that  biospace  filling  processes  are  more  advanced 
and  that  the  formation  is  structurally  more  stable. 

As  already  stated,  the  three  types  that  complete  the 
structural  typology  of  the  first  set  are  each  represented 
by  only  one  observation.  They  illustrate  structural  condi- 
tions that  are  not  very  frequent  but  clearly  distinct  from 
those  examined  so  far.  In  particular,  the  third  and  fovirth 
type  distinctly  display  some  possible  early  stages  of  the 
regeneration  process,  since  two  clearly  different  genera- 
tions coexist  at  the  same  time.  A  high  proportion  of  trees 
over  220  years  of  age  can  be  observed  in  both  types,  while 
the  remaining  individuals  fall  within  yoimger  classes.  The 
structure  of  the  third  type  can  be  easily  explained  consider- 
ing that  the  plot  is  located  at  the  treeline  (2,260  m).  This  is 
an  environment  where  the  regeneration  process  occurs  only 
occasionally  or  it  may  be  entirely  missing  for  a  long  time. 
The  latter  seems  to  have  happened  in  this  case. 

In  the  fourth  type,  the  large  number  of  trees  per  hec- 
tare (625)  suggests  that  the  future  structure  in  this  stand 
should  be  similar  to  the  first  two  types  described. 

The  last  type,  as  already  pointed  out,  represents  a 
rather  rare  situation. 

In  the  first  type  of  the  second  set,  the  chronological 
analysis  suggests  that  the  stand  originated  from  a  grazed 
forest.  Spruce,  as  well  as  stone  pine,  is  represented  in  all 
classes,  which  means  that  there  was  a  continuous  inva- 
sion of  these  two  species  into  the  formation. 

The  second  type  is  characterized  by  formations  that  are 
open  grown,  permitting  new  spruce  and  stone  pine  to  con- 
tinuously invade  the  stand.  However,  larch  regeneration 
ended  about  80  years  ago. 

In  the  last  type,  regeneration  occurred  rapidly  and  a 
sufficiently  closed  stand  established  in  a  short  time.  After 
that,  regeneration  slowed  down,  but  competition  among 
individuals  increased.  That  caused  the  diameter  differen- 
tiation mentioned  earlier. 


DISCUSSION 

For  the  first  two  sets  of  stand  structure,  homogeneous 
structural  types  can  be  interpreted  from  the  chronological 
point  of  view.  The  third  set  was  different. 

In  the  first  type,  physical  elements  show  a  structure 
characterized  by  an  intermediate  layer  that  is  sufficiently 
differentiated  by  competition  and  is  mixed  with  remark- 
ably large-size  trees  from  a  previous  stand.  Taking  into 
consideration  age  distribution,  few  trees  older  than  120 


CONCLUSIONS 

From  the  methodological  point  of  view,  the  use  of  dis- 
tance measures  has  made  it  possible  to  point  out  a  struc- 
tural typology  in  stands  where  stone  pine,  or  spruce  with 
stone  pine,  are  clearly  predominant.  These  results  were 
obtained  using  physical  structure  elements  as  discrimi- 
nating factors.  Other  results — more  difficult  to  interpret — 
were  obtained  that  take  into  consideration  both  physical 
and  chronological  elements. 


236 


10  15  20  25  30  35  40  45  50  55  60  65  70  75 
diameter  classes  (cm) 


T  r  Y  T 

21-40      M-W      fil-flO     SMCn    101.12U    I2M  40    141-160    ICiMMU    IKi-20Q   2UI-220  >221 


chronological  age  classes 


I     I  P.  cembra  IB  L.  decidua        P.  abies 


Figure  8— Structural  type  2.2;  including  plots  5,  34,  36, 
and  39. 


237 


2> 


238 


The  structural  typology  defined,  though  rather  variable, 
shows  that  the  regeneration  process  started  about  one 
century  ago,  and  it  sometimes  developed  in  a  gradual  and 
continuous  way  and  at  other  times  more  quickly  and  mas- 
sively. Generally,  it  is  possible  to  observe  good  structures 
in  physical  and  chronological  elements.  These  will  prob- 
ably provide  these  stands  with  sufficient  stability. 

REFERENCES 

Contini,  L.;  Lavarelo,  Y.  1981.  Le  pin  cembro:  repartition, 
ecologie  et  croissance.  CNRF,  Nancy.  254  p. 

Contini,  L.;  Lavarelo,  Y.  1982.  Le  pin  cembro:  vegetation, 
ecologie,  Sylviciilt\ire  et  production.  INRA,  Paris.  197  p. 

Del  Favero,  R.;  De  Mas,  G.;  Lasen,  C;  Paiero,  P.  1985.  II 
pino  cembro  nel  Veneto.  Assessorato  Agric.  e  For.,  Dip. 
per  le  foreste  e  I'economia  montana,  Venezia.  85  p. 

FiUpello,  S.;  Sartori,  F.;  Vittadini,  M.  1976.  Le  associazioni 
del  cembro  sul  versante  meridionale  dell'arco  alpino.  Atti 
Istituto  Botanica  Universita  di  Pavia.  11:  21-104. 

Filipello,  S.;  Sartori,  F.;  Vittadini,  M.  1980-1981.  Le  associ- 
azioni di  cembro  nel  versante  meridionale  dell'arco 
alpino.  2:  La  vegetazione:  aspetti  forestali.  Atti  Istituto 
Botanica  Universita  di  Pavia.  14:  1-48. 

Fliri,  F.  1975.  Das  Klima  der  Alpen  im  Raimie  von  Tirol. 
Universitaets-Verlag  Wagner,  Innsbruck-Mxmchen. 
454  p. 

Franceschetti,  G.;  Provasi,  C.  1978.  La  cluster  analysis 
quale  strumento  di  zonizzazione  nella  pianificazione. 
In:  Metodologia  per  la  formazione  dei  piani  zonali 


agricoH  nella  Regione  Veneto.  Istituto  Estimo  Rxirale 

e  Contabilita,  Universite  di  Padova:  32-90. 
I.  S.  A.  F.  A.  1984.  Inventaiio  forestale  nazionale.  Tavole 

di  cubatura  a  doppia  entrata.  Trento.  Ill  p. 
Kuoch,  R.  1972.  Zur  Struktur  und  Behanlimg  von 

Subalpinen  Fichtenwaeldem.  Schw.  Zeit.  fur  Forst. 

123(2):  77-89. 

Mayer,  H.;  Ott,  E.  1991.  Gebirgswaldbau  Schutzwald- 
pflege.  2  Auflage,  Gustav  Fischer  Verlag  Stuttgart- 
New  York. 

Piussi,  P.;  Schneider,  A.  1985.  I  limiti  superiori  del  bosco 
e  degli  alberi  in  Val  di  Vizze  (prov.  di  Bolzano).  Annali 
Acc.  It.  Sc.  Forestali:  121-150. 

Rachoy,  W.  1976.  Waldbauliche  Stniktunmterschungen 
in  subalpinen  Zirbenwaeldem.  Invited  Papers,  Con- 
gress Group  1,  XVI.  lUFRO-World-Congress,  Oslo  1976: 
183-202. 

Schiechtl,  H.  M.;  Stem,  R.  1975.  Die  Zirbe  iPinus  cemhra 
L.)  in  den  Ostalpen.  I  Teil,  Angew.  Pflanzensoz.  22: 
1-83. 

Schiechtl,  H.  M.;  Stem,  R.  1979.  Die  Zirbe  {Pinus  cemhra 
L.)  in  den  Ostalpen.  II  Teil,  Angew.  Pflanzensoz.  24: 
1-79. 

Schiechtl,  H.  M.;  Stem,  R.  1983.  Die  Zirbe  {Pinus  cembra 
L.)  in  den  Ostalpen.  Ill  Teil,  Angew.  Pflanzensoz.  27: 
1-110. 

Schiechtl,  H.  M.;  Stem,  R.  1984.  Die  Zirbe  (Pinus  cembra 
L.)  in  den  Ostalpen.  IV  Teil,  Angew.  Pflanzensoz.  28: 
1-99. 

Stem,  R.;  Helm,  G.  1979.  Alter  und  Struktur  von 
Zirbenwaldem.  Allgem.  Forstr.  90:  194-198. 


239 


ROLE  OF  PINUS  PUMILA  IN  PRIMARY 
SUCCESSION  ON  THE  LAVA  FLOWS  OF 
VOLCANOES  OF  KAMCHATKA 

S.  Yu.  Grishin 


Abstract— The  succession  of  the  subalpine  vegetation  was  studied 
on  the  lava  flows  of  the  Central  Kamchatka.  The  rate  of  succes- 
sion depends  on  type  of  lava  and  the  formation  of  friable  substra- 
tum  on  lava  surface  following  ash  falls  and  other  causes.  Stone 
pine  (Pinus  pumila)  is  not  a  pioneer  on  the  original  lava  surface. 
Rather,  its  main  role  is  determined  during  the  first  half  of  succes- 
sion (0  to  1,000  years).  During  this  period  it  forms  closed  commu- 
nities and  soil  cover,  favoring  the  establishment  of  other  dominant 
species  in  the  subalpine  zone.  During  the  second  half  of  succession 
(1,000  to  2,000  years)  P.  pumila  decreases  to  the  cover  level  typi- 
cal for  climax  vegetation.  Thxis  the  role  of  P.  pumila  is  unique;  it 
restores  vegetation  cover  on  juvenile  substrate. 


Many  thousand  square  kilometers  of  Kamchatka's  terri- 
tory are  covered  by  layers  of  lava  flows.  Most  of  the  terri- 
tory has  subalpine  vegetation.  Under  these  conditions  the 
subalpine  stone  pine  (Pinus  pumila)  (nomenclature  of  all 
species  follows  Czerepanov  1981)  is  the  most  important  spe- 
cies forming  plant  cover  on  juvenile  substrate.  This  spe- 
cies usually  forms  dense  commimities  in  the  subalpine  belt 


vole.  Shiveluch  # 


Figure  1 — Location  of  study  area. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

S.  Yu.  Grishin  is  Research  Scientist  at  the  Institute  of  Biology  and  Pedol- 
ogy of  Far  East  Branch  of  Russian  Academy  of  Sciences,  Vladivo8tok-22 
690022,  Russia. 


Figure  2— Tolbachik  volcano  (altitude  3,682  m). 


in  northeastern  Asia.  At  Kamchatka,  p-la  P.  pumila  may 
be  found  from  the  seashore  to  the  mountain  slopes  up  to 
1,400  m  above  sea  level.  It  covers  42  percent  of  the  forest- 
covered  territory. 

This  study  was  conducted  in  the  Central  Kamchatka, 
near  Tolbachik  volcano  (figs.  1  and  2).  During  the  Holocene, 
lava  flows  formed  a  lava  plateau  with  an  area  of  875  km'^. 
The  lava  flows  of  Kluchevskoy  volcano  were  also  investi- 
gated. The  age  of  the  flows  was  determined  by  the  tephra- 
chronology  method,  based  on  radiocarbon  dating  (Braitseva 
and  others  1981).  With  this  method  we  could  distinguish 
the  stages  of  succession  and  estimate  their  duration 
(Grishin  1992). 

The  climatic  conditions  of  the  Tolbachik  area  reveal  a  con- 
tinental pattern,  similar  to  the  climate  of  Middle  Siberia. 
The  mountain  vegetation  is  composed  by  forests  of  spruce 
(Picea  ajanensis)  and  larch  (Larix  kamtschatica)  up  to  an 
altitude  of  500  m,  while  birch  (Betula  ermanii)  occurs  up  to 
900  m.  The  subalpine  zone  is  situated  at  the  interval  of 
800  to  1,100  m.  It  Dhas  a  complex  structure:  forest  islands 
of  birch,  open  woodlands  of  larch,  vast  covers  of  stone  pine 
and  alder  (Duschekia  kamtschatica)  krummholz,  patches 
of  subalpine  meadows  and  alpine-tj^je  communities.  The 
alpine  belt  extends  to  2,000  m  altitude,  and  the  snowline 
lies  near  2,500  m  (Grishin  1988a,b). 

In  the  summer  of  1975  in  the  central  part  of  the  plateau 
there  was  a  major  eruption.  As  a  result  of  this,  lava  flows 
with  an  area  of  about  9  km^  were  effused  and  tephra  (vol- 
canic ash,  s£ind,  and  scoria)  was  dispersed  on  470  km^.  This 


240 


Figure  3 — Volcanic  plateau  of  the  Tolbachik  area: 
lava  flows  covered  by  tephra  and  remnants  of  larch 
trees  and  Pinus  pumila  krummholz. 

layer  has  a  depth  of  more  than  0.1  m  (fig.  3).  The  vegeta- 
tion was  completely  destroyed  on  400  km^  (Fedotov  1984). 

ROLE  IN  SUCCESSION 

The  investigation  of  the  lava  flows  (10  to  50  years  old) 
showed  that  P.  pumila  is  not  a  pioneer  species  actively  es- 
tablishing on  the  juvenile  substrata  of  pahoehoe  (ropy  lava), 
aa  (scoria  lava),  and  tephra.  Herbaceous  plants  (Chamerion 
angustifolium,  Poa  malacantha,  Leymus  ajanensis,  and 
others),  mosses  (Polytrichum  spp.),  and  lichens  (Stereocau- 
lon  vesuvianum)  establish  most  actively  initially.  Woody 
plants  such  as  poplar  {Populus  suaveolens),  some  willows, 
and  very  occasionally  larch  and  stone  birch  establish  less 
readily — all  the  plants  settle  on  the  friable  substratimi 
(tephra,  products  of  weathering  of  lava,  and  others).  Seed- 
lings of  P.  pumila  were  rather  rare,  appearing  a  few  years 


Figure  4 — Seedling  of  Pinus  pumila  on  aa  lava. 


Figure  5 — Lava  flow  about  500  years  old  coverea 
by  tephra  to  a  depth  of  about  40  cm. 


after  the  eruption  (fig.  4).  The  difficulties  in  its  distribution 
are  because  its  seeds  are  dispersed  by  animals  and  seed 
sources  5  to  10  km  away.  Five  himdred  years  later,  scat- 
tered herbaceous  and  shrubby  vegetation  is  formed  on  lava 
surface  with  single  (about  10  trees/ha),  extremely  stimted 
(3  to  5  m  height)  larch  trees.  Low  thickets  of  P.  pumila 
(0.3  to  0.5  m  height)  dominate  the  vegetation,  covering 
about  30  percent  of  the  surface  (fig.  5).  Pinus  pumila  fills 
all  potential  habitats  with  friable  substratimi  and  forces 
out  the  herbaceous  pioneer  species.  Under  the  tephra  de- 
posit on  blocks  and  plates  of  lava,  a  thin  layer  (1  to  3  cm)  of 
primitive  soil  develops.  Pinus  pumila  is  rooted  in  crevices 
between  blocks  on  concentrated  finable  substratum  enriched 
with  organic  matter  (fig.  6). 

On  the  1,000-year-old  lava  flows,  more  closed  vegetation 
can  be  observed.  Pinus  pumila  is  the  dominant  species  and 
covers  60  to  90  percent  of  the  surface  (fig.  7).  Its  height 


Figure  6 — The  remnants  of  stunted  larch  and 
branches  of  Pinus  pumila  between  blocks  of  lava. 


241 


Figure  7 — The  flow  of  aa  lava  about  1 ,000  years  old. 


reaches  1  to  1.5  m.  The  characteristic  feature  of  this  stage 
of  succession  is  the  appearance  of  dominants  of  new  commu- 
nities. The  most  important  plant  in  the  new  communities 
is  alder  krummholz.  It  settles  on  the  substratum  prepared 
by  P.  pumila,  which  eventually  disappears  from  the  slopes 
of  the  lava  ridge  and  occurs  only  in  the  flat  bottoms  between 
ridges  (fig.  8).  Together  with  stone  pine,  slowly  growing 
small  larch  trees  (a  few  dozen  trees/ha)  are  found.  The 
abimdant  leaf  fall  of  the  alder  and  of  the  undergrowth  plants 
(Calamagrostis  langsdorffii,  and  others)  considerably  accel- 
erates the  rate  of  succession  and  prepares  the  substratvmi 
for  stone  birch — a  very  important  dominant  tree  in  the  sub- 
alpine  forest.  The  depth  of  the  soil  on  different  lava  flows 
is  5  to  15  cm. 

During  the  next  500  years  (1,000  to  1,500  years  old)  the  rate 
of  succession  increases  noticeably,  leading  to  differentiation 
of  vegetation  in  the  subcdpine  zone.  On  the  1,500-year-old 


lava  flows,  typical  subalpine  communities  can  be  distin- 
guished. In  addition  to  the  dominant  P.  pumila  thickets, 
the  single  clumps  of  alder,  and  the  open  woodlands  of  larch, 
communities  of  stone  birch,  covers  of  alder,  subalpine  mead- 
ows, and  fragments  of  alpine  heath  appear.  At  this  point, 
the  coverage  of  P.  pumila  decreases,  occupying  its  own 
typical  habitats  such  as  rock  crests,  concave  hollows  with 
frozen  soils,  and  wind-exposed  sites.  The  thickness  of  the 
soil  profile  increases  to  20  cm,  and  the  leveling  of  volcano- 
genic  mi crorelief  begins  to  occur.  Both  coenotical  and  ver- 
tical differentiation  of  vegetation  takes  place.  During  the 
earlier  stage  of  succession,  P.  pumila  dominated  at  altitudes 
fi'om  200  to  1,000  m,  but  to  this  stage  the  forest  belt  below 
800  m  is  formed,  composed  of  larch  and  stone  birch. 

On  the  2,000-  and  2,500-year-old  lava  flows  the  vegeta- 
tion was  destroyed  completely  by  the  ash  fall  of  1975  (fig.  9). 
Its  remnants  give  evidence  that  vegetation  cover  had  been 
approaching  the  climax  stage.  It  had  mature  commxmities 
with  well- developed  trees  and  krvunmholz,  the  size  and 
cover  of  which  were  related  to  their  altitudinal  position. 
For  example,  at  one  of  the  flows  at  the  altitude  of  700  m 
a  well-developed  birch-larch  subalpine  forest  with  some 
spruce  was  located.  The  birch  trees  had  a  height  of  more 
than  10  m  and  a  diameter  at  breast  height  of  25  to  40  cm. 
The  larch  trees  were  about  20  m  and  30  to  40  cm,  respec- 
tively. The  alder  krummholz  under  the  forest  layer  had  a 
height  of  2.5  to  3  m,  and  P.  pumila  heights  were  1.5  to  2  m. 

These  examples  show  that  under  such  conditions,  begin- 
ning at  the  age  of  1,500  years,  the  vegetation  tends  to  ap- 
proach the  climatic  climax  similar  to  the  neighboring  locah- 
ties.  Many  criteria  prove  it. 

The  peculiarities  of  succession  are  determined  mainly 
by  the  pattern  of  lava  cover  and  the  accumulation  of  fiiable 
substratvmi  on  the  lava  surface.  On  pahoehoe  lava,  estab- 
lishment of  plants  on  the  smooth  monolith  surface  is  an 
extremely  slow  process.  After  the  first  1,000  years  of  such 
lithosere,  the  lava  is  commonly  covered  only  by  crustaceous 
lichens.  Higher  plants  settle  on  friable  substratvun  in  cracks 
and  cavities,  and  that  is  in  fact  the  beginning  of  psammo- 
sere  (fig.  10).  Even  if  these  plants  are  rooted  in  lava,  they 


6 


Figure  8 — The  distribution  of  vegetation  across  aa  lava  flow  about  1 ,000  years  old.  1 — alder  krummholz; 
2 — shrubs  of  Pinus  pumila;  3 — petrophilous  apeu  of  groups  in  the  lava  outcrops;  4 — bunches  of  grasses; 
5 — lava  matter;  6 — tephra  of  1975;  7 — buried  snow  patches. 


242 


Figure  9 — Timberline  ecotone  vegetation  killed  by  ash 
fall  on  a  lava  flow  that  was  about  2,000  years  old. 


Bezymyanny  were  also  deposited.  The  negative  influence 
of  major  ash  falls  on  vegetation  was  shown  by  the  1975 
eruption.  Pinus  pumila  krummholz  and  subalpine  birch 
forests  were  killed  under  the  tephra  deposit  of  more  than 
20  cm,  alder  krummholz  imder  more  than  30  cm,  and  larch 
forests  imder  more  than  40  cm.  Smaller  amounts  of  tephra 
(10  to  15  cm)  lead  to  the  death  of  vegetation  in  subalpine 
meadows  and  alpine  heaths.  After  moderate  ash  fall,  a  sec- 
ondary succession  is  possible,  and  thus  lithosere  may  be 
transformed  to  psammosere. 

There  are  other  essential  factors  of  succession  such  as  cH- 
matic  fluctuations  of  forest  and  subalpine  vegetation  at  their 
upper  limit  and  forest  fires.  Fires  may  originate  from  vol- 
canic eruption.  Communities  of  coniferous  trees  (larch, 
spruce)  and  especially  krimimholz  (P.  pumila)  with  dry 
lichens  and  dwarf  shrubs  from  Ericacaea  are  only  slightly 
protected  from  fires. 

Thus,  the  role  of  P.  pumila  in  the  succession  on  the  lava 
flows  of  the  Tolbachik  area  is  different  at  diff'erent  stages 
of  succession.  Because  of  its  ecological  and  biological  char- 
acteristics, it  is  not  a  pioneer  at  the  initial  stage.  Pinus 
pumila  plays  its  main  role  during  the  first  half  of  succes- 
sion (up  to  approximately  1,000  years).  It  forms  closed 
commimities  and  soil  cover,  and  favors  the  establishment 
of  dominant  species  in  new  communities  of  the  subalpine 
zone.  During  the  second  half  of  succession,  plant  cover  of 
P.  pumila  gradually  decreases  to  the  level  typical  of  climax 
vegetation  (fig.  11). 

CONCLUDING  REMARKS 

On  the  whole,  the  role  of  P.  pumila  in  succession  is  unique: 
It  restores  vegetation  cover  on  juvenile  substrate  unfavor- 
able for  the  establishment  and  development  of  woody  and 
sluiibby  species.  This  feature  of  P.  pumila  is  realized  espe- 
cially in  the  regions  with  continental  climate.  In  places  with 
oceanic  cUmate,  for  example  on  the  slopes  of  the  Kluchevskoy 
volcano,  the  coniferous  dominants  P.  pumila  and  Larix 


Figure  10 — Suppressed  Pinus  pumila  and  stunted 
trees  of  larch  on  pahoehoe  lava  about  1 ,000  years  old. 


cannot  populate  all  the  surface  of  the  pahoehoe,  and  the 
next  stages  of  succession  are  dependent  on  the  space  corre- 
lation between  lithosere  and  psammosere.  On  aa  lava  this 
correlation  changes  to  psammosere  more  qvdckly,  and  that 
accelerates  succession.  The  final  stages  of  succession  on 
different  types  of  lava  are  probably  similar  because  a  thick 
soil-tephra  profile  is  gradually  formed,  smoothing  out  the 
irregularities  of  lava. 

Another  factor  important  to  the  succession  rate  is  climate. 
In  the  area  studied,  the  duration  of  succession  is  two  to  three 
times  longer  than  in  the  oceanic  climate  of  the  Japanese 
subtropics,  where  it  continues  for  only  700  years  (Tagawa 
1964). 

A  very  important  factor  following  volcanism  is  ash  fall 
damage  to  vegetation.  For  example,  in  the  Tolbachik  area, 
about  10  eruptions  took  place  during  the  last  2,000  years. 
Four  of  them  were  similar  to  the  eruptions  of  1975;  in  that 
period  the  tephra  of  volcanoes  Shiveluch,  Kluchevskoy,  and 


1,000  1,500 

Age  (yr) 


Figure  11 — Generalized  chronocline  of  succession 
on  lava  flows  of  the  Tolbachik  are^.  1 — pioneer 
unclosed  groups  of  grasses,  mosses,  lichens; 
2 — communities  of  Pinus  pumila;  3 — woodlands 
of  larch;  A — communities  of  alder  krummholz; 
5 — birch  forests;  6 — subalpine  meadows. 


2,000 


243 


kamtschatica  are  replaced  by  deciduous  species  such  as 
Populus  suaveolens,  Duschekia  kamtschatica,  and  Salix  spp. 
Nevertheless,  P.  pumila  is  widely  distributed  in  the  mo;in- 
tains  of  the  volcanic  island  arc  of  the  northwest  Pacific 
(Kamchatka,  Kurile,  and  Japan  islands).  Pinus  pumila 
may  probably  be  used  to  restore  vegetation  on  eroded  and 
rocky  places  in  other  high-mountain  regions  of  the  North- 
ern Hemisphere. 

REFERENCES 

Braitseva,  O.  A.;  Melekescev,  I.  V.;  Ponomariova,  V.  V.;  and 
others.  1981.  Tephrachronological  and  geochronological 
investigation  of  Tolbachik  regional  zone  of  scoria  cones. 
Volcanology  and  Seismology.  (2):  14-28.  [In  Russian]. 

Czerepanov,  S.  K.  1981.  Plantae  vasculares  URSS. 
Leningrad:  Nauka.  510  p. 

Fedotov,  S.  A.,  ed.  1984.  A  major  fissure  Tolbachik  erup- 
tion (1975-1976,  Kamchatka).  Moscow:  Nauka.  638  p. 
[In  Russian]. 


Grishin,  S.  Yu.  1988a.  The  upper  timberline  at 
Kluchevsakaia  volcano  group.  In:  Kharkevicz  S.  S., 
resp.  ed.  Plant  cover  of  high-mountain  ecosystems  of 
the  USSR.  Vladivostok:  Far  East  Branch  of  Academy 
of  Sciences  of  the  USSR:  193-201.  [In  Russian]. 

Grishin,  S.  Yu.  1988b.  Structure  of  timberline  ecotone 
vegetation  on  Mt.  Dalnaia  Ploskaia  (Kamchatka).  In: 
Komarovskie  chtenia,  XXXV.  Vladivostok:  Far  East 
Branch  of  Academy  of  Sciences  of  the  USSR:  159-175. 
[In  Russian]. 

Grishin,  S.  Yu.  1992.  The  successions  of  subalpine  vege- 
tation on  the  lava  flows  of  Tolbachik  area.  Botanical 
Journal  [of  Russian  Botanical  Society].  77(1):  92-100. 
[In  Russian]. 

Tagawa,  H.  1964.  A  study  of  the  volcanic  vegetation  in 
Sakurajima,  South- West  Japan.  1.  Dynamics  of  vegeta- 
tion. JVIem.  Fac.  Sci.  Kysushu  Univ.  Ser.  E(Bot.).  3(3-4): 
165-228. 


244 


DECLINE  OF  WHITEBARK  PINE  IN 
THE  BOB  MARSHALL  WILDERNESS 
COMPLEX  OF  MONTANA,  U.S.A. 

Robert  E.  Keane 
Penelope  Morgan 


Abstract— Populations  of  whitebark  pine  {Pinus  albicaulis)  in 
the  Northern  Rocky  Movintains,  USA,  are  being  reduced  at 
alarming  rates  due  to  combined  or  individual  effects  of  blister 
rust  (Cronartium  ribicola),  movmtain  pine  beetle  (Dendroctonus 
ponderosae),  and  advancing  succession  resulting  from  fire  sup- 
pression. Results  of  an  extensive  field  survey  of  various  white- 
bark  pine  communities  were  used  to  evaluate  past  and  cvirrent 
whitebark  pine  population  levels.  The  ecological  process  model 
FIRESUM  was  then  used  to  simulate  forest  succession  with  ef- 
fects of  blister  rust  and  long-term  fire  suppression.  Results  indi- 
cate whitebark  pine  population  levels  are  significantly  decreas- 
ing, mostly  as  a  result  of  blister  rust,  but  decreases  may  be 
mitigated  by  the  reintroduction  of  fire. 


Whitebark  pine  (Pinus  albicaulis)  is  a  common  tree  spe- 
cies of  Northern  Rocky  Mountain  upper  subalpine  forests 
and  timberlines.  In  Montana,  Idaho,  and  northwestern 
Wyoming,  USA,  whitebark  pine  is  an  important  compo- 
nent of  about  10-15  percent  of  the  forested  landscape.  Its 
slow  growth,  modest  stature,  and  inaccessible  habitats 
generally  make  it  a  low  value  commercial  timber  species. 
However,  its  cones  are  highly  valued  by  many  species  of 
wildlife  as  a  source  of  food.  Animals  that  utUize  white- 
bark pine  cone  crops  include  black  and  grizzly  bears 
{Ursus  americanus  and  U.  arctos  horribilis),  red  squirrels 
{Tamiasciurus  hudsonicus)  (Femer  1974),  and  the  Clark's 
nutcracker  (Nucifraga  columbiana)  (Kendall  1980; 
Mattson  and  Reinhart  1986).  The  nutcracker  plays  an 
important  mutualistic  role  in  whitebark  pine  regeneration 
because  it  is  essentially  the  only  dispersal  vector  for 
whitebark  pine  seed  (Tomback  1982;  Tomback  and  others 
1990).  Whitebark  pine  is  also  important  for  snow  reten- 
tion and  watershed  protection  in  high-elevation  areas 
where  no  other  species  can  become  established  (Hann 
1990). 

Whitebark  pine  populations  have  been  observed  to  be 
dechning  in  parts  of  the  Northern  Rocky  Mountains  ( Amo 
1986;  Ciesla  and  Fumiss  1986;  Kendall  and  Arno  1990; 
Moore  1984).  The  cause  of  the  decline  has  been  mainly 
attributed  to  mountain  pine  beetle  (Dendroctonous 
ponderosae),  successional  advancement,  and  white  pine 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Robert  E.  Keane  is  Research  Ecologist,  Intermountain  Fire  Sciences 
Laboratory,  Intermountain  Research  Station,  Forest  Service,  U.S. 
Department  of  Agriculture,  P.O.  Box  8089,  Missoula,  MT,  USA  59801; 
Penelope  Morgan  is  Professor  of  Fire  Ecology,  Department  of  Forest 
Resources,  University  of  Idaho,  Moscow,  ID,  USA  83843. 


bhster  rust  (Cronartium  ribicola).  Moimtain  pine  beetle 
epidemics  killed  many  whitebark  pine  trees  during  the 
early  1900's  (Amo  1970;  Amo  and  Hoff  1989).  Extensive 
successional  replacement  of  whitebark  pine  by  subalpine 
fir  (Abies  lasiocarpa)  and  Engelmann  spruce  (Picea  engel- 
mannii)  may  be  a  direct  result  of  more  than  a  half  centiuy 
of  fire  suppression  (Amo  1986).  The  most  damaging 
agent,  bhster  rust,  is  an  introduced  disease  that  is  espe- 
cially devastating  to  whitebark  pine  (Amo  and  HofF  1989; 
BedweU  and  Childs  1943;  Bingham  1972;  Hoff  and  others 
1980).  Unfortunately,  the  extent  and  severity  of  white- 
bark pine  decUne  is  unknown  in  the  Northern  Rocky 
Mountains  because  documentation  has  been  mostly  from 
casual  observations  rather  than  scientific  investigation. 

This  study  was  initiated  to  determine  historical,  cur- 
rent, and  future  whitebark  pine  population  levels  in  a  por- 
tion of  the  Northem  Rocky  Mountains  known  as  the  Bob 
Marshall  Wilderness  Complex.  Whitebark  pine  communi- 
ties were  intensively  inventoried  throughout  this  study 
area.  Community  structure  and  age  information  were 
used  to  reconstruct  historical  community  compositions. 
An  ecosystem  process  model  was  used  with  sampled  data 
to  predict  the  future  of  these  forests  under  foiir  scenarios. 

STUDY  AREA 

The  Bob  Marshall  Wilderness  Complex  (BMWC)  is  a  re- 
mote 600,000-hectare  preserve  in  north westem  Montana, 
USA,  composed  of  the  Great  Bear,  Bob  Marshall,  and 
Scapegoat  WUdemesses  (fig.  1).  This  large  area  consists 
of  mountainous  terrain  dissected  by  large  river  drainages. 
Parent  material  is  mostly  qu£irtzite  and  argillite  with  al- 
temating  layers  of  limestone.  The  Continental  Divide 
runs  through  the  wilderness  creating  a  unique  blend  of 
climates  and  plant  communities.  Climate  west  of  the  Di- 
vide is  mostly  maritime-continental  with  cool  wet  winters 
and  warm  dry  summers.  The  east  side  climate  is  mostly 
continental  with  cold  dry  winters  and  warm  dry  summers. 

Whitebark  pine  is  a  forest  component  above  approxi- 
mately 1,800  m  in  the  study  area.  It  has  the  potential  to 
dominate  20-30  percent  of  the  landscape  v^thin  the  entire 
study  area  (Bain  1989).  Whitebark  pine  is  serai  to  subal- 
pine fir  and  Engelmann  spmce  in  most  of  the  area,  but  it 
can  form  climax  stands  on  high,  dry  ridge  and  mountain 
tops. 

Fire  was  a  dominant  process  on  the  BMWC  landscape. 
Ayers  (1901)  estimated  40  percent  of  BMWC's  whitebark 
pine  forests  were  burned  between  1858  and  1898.  Large, 
stand-replacement  fires  are  typical  in  the  study  area, 
especially  in  the  whitebark  pine  zone  (Losensky  1990). 


245 


10   0  10  20  30  40     .  , 

1-1  i-i  i-i  i-j  i_i  I  — I  h  i  —I  Statute  Miles 

10  0  10  20  30  40  50 

M  M  H  M  H I  -I  I  I  I        ^3  Kilometers 

Figure  1 — Bob  Marshall  Wilderness  Complex  in  the  west-central  portion  of  Montana,  USA,  with  loca- 
tions of  all  whitebark  pine  sample  sites  across  the  complex  shown  by  the  "stars." 


246 


The  great  distances  that  nutcrackers  transport  seed  allow 
whitebark  pine  a  competitive  advantage  in  colonizing  vast 
burned  areas  (Tomback  1990).  Clark's  nutcrackers  prefer 
open  areas  to  cache  whitebark  pine  seeds  (Sund  and  oth- 
ers 1991).  Some  whitebark  pine  stands  contain  evidence 
of  low-severity,  frequent  smface  fires  (Gabriel  1976). 
These  fires  kill  many  subalpine  fir  trees  that  compete 
with  the  more  fire-resistant  whitebark  pine  (Amo  1986). 


and  also  recorded  for  the  entire  stand.  Blister  rust  sever- 
ity was  evaluated  for  each  tree  fi-om  (1)  nxunber  of  visible 
cankers  per  tree,  (2)  nimiber  of  infected  trees  within  the 
macroplot,  and  (3)  amount  of  tree  foliage  killed  by  the 
rust.  Causes  of  mortality  were  estimated  for  dead  trees 
when  evidence  existed.  When  possible,  snags  and  dead 
down  trees  were  identified  to  species  for  historical  stand 
reconstruction. 


MODEL  DESCRIPTION 

FIRESUM  is  a  FORTRAN  77  computer  model  used  to 
simulate  important  ecological  processes  in  forested  eco- 
systems. A  complete  description  of  FIRESUM  is  given 
in  Keane  and  others  (1989).  In  general,  effects  of  shading, 
crowding,  water  availability,  and  cHmate  on  tree  growth, 
regeneration,  and  mortality  are  simulated  from  stand 
structure,  weather,  and  soils  information.  Trees  die  either 
from  stress-induced  mortahty  evaluated  fi-om  the  annual 
diameter  increment,  or  from  natural  cavises  modeled  from 
stochastic  functions.  Fires  can  kill  trees,  and  they  also  re- 
duce fuel  loadings  and  duff/Utter  depths.  Woody  fuels  are 
modeled  empirically  and  used  to  compute  fire  intensity 
and  spread  rates.  Duff  and  litter  depths  are  simulated 
dynamically  from  Htterfall  and  decomposition  rates  then 
used  to  reduce  regeneration  rates  or  promote  fire  spread. 

Regeneration  of  wind-dispersed  tree  species  is  modeled 
stochastically  in  FIRESUM  from  the  processes  mentioned 
above  and  also  fi-om  cone  crop  fi-equency,  distance  from 
seed  source,  and  stand  composition.  However,  since  white- 
bark pine  seeds  are  dispersed  by  the  Clark's  nutcracker, 
additional  simulation  algorithms  were  added  to  FIRESUM 
to  account  for  Clark's  nutcracker  population  levels,  site 
suitability  for  seed  caching,  and  cone  availabihty.  These 
algorithms  are  presented  in  Keane  and  others  (1990a,b) 
and  provide  the  means  for  modeling  whitebark  pine  cone 
crops. 

FIELD  METHODS 

Stand  structure,  fire  history,  fuels,  site,  and  plant  com- 
munity information  were  sampled  in  various  whitebark 
pine  stands  across  the  entire  BMWC  (Keane  and  others, 
in  preparation).  Trees  were  intensively  sampled  for  age, 
size,  and  vigor  within  a  400-m^  circular  macroplot.  Tree 
fire  scars  and  tree  ages  were  used  to  obtain  a  fire  history 
(Amo  and  Sneck  1977).  Down,  dead  woody  fuels  (twigs, 
branches,  and  logs)  were  inventoried  around  the 
macroplot  using  transect  techniques  (Brown  1974). 
ECODATA  methods  (Hann  and  others  1988;  Keane  and 
others  1990c)  were  used  to  record  site  information  such  as 
elevation,  aspect,  slope,  and  vascular  plant  community 
composition  by  species.  Macroplots  were  located  in  a  rep- 
resentative portion  of  whitebark  pine  stands  and  then  ac- 
curately located  using  Global  Positioning  Systems  (Hum 
1989)  so  the  stand  coiild  be  recognized  from  satellite  im- 
agery (Keane  and  others,  in  preparation). 

Blister  mst  and  mountain  pine  beetle  evidence  was  re- 
corded for  each  whitebark  pine  tree  within  the  macroplot. 


SIMULATION  METHODS 

FIRESUM  was  used  to  investigate  effects  of  four  eco- 
logical disturbance  scenarios  on  tree  species  at  the  Cliff 
Mountain  site,  located  in  the  central  portion  of  the  BMWC 
at  2,100  m  elevation.  Tree  species  input  parameters  to 
FIRESUM  were  taken  from  previous  FIRESUM  simxila- 
tion  exercises  (Keane  and  others  1990a).  Chmate  inputs 
were  derived  from  weather  data  taken  at  a  station  just 
outside  the  BMWC  boundary  in  the  town  of  Seeley  Lake. 
The  model  was  calibrated  with  field  data  sampled  from 
adjacent  distvu-bed  and  matxire  whitebark  pine  com- 
munities on  the  Shale  Mountain  site,  a  sampled  stand  at 
2,200  m  on  northwest  aspect  in  south-central  BMWC. 
The  mature  commimity  tree  data  were  used  as  initial 
stand  condition  inputs,  and  fire  history  of  the  disturbed 
community  was  used  for  disturbance  input  parameters  to 
FIRESUM.  Subsequent  simulation  results  were  com- 
pared to  actual  data  from  the  disturbed  community 
(Keane  and  others  1990a,b).  As  a  result  of  these  tests, 
model  parameters  and  equations  were  adjusted  to  more 
closely  approximate  observed  succession  in  these  forests. 

The  following  four  scenarios  were  simulated  in 
FIRESUM  for  the  Cliff  Mountain  site  over  a  500-year 
period  after  parameter  adjustment: 

1.  Fires  at  stochastic  intervals  averaging  150  years  and 
no  bhster  rust  or  beetle  infestations. 

2.  No  fires  (complete  fire  suppression)  with  no  blister 
rust  or  beetle  infestations. 

3.  Blister  mst  infestation  at  year  100  and  no  fires. 

4.  Blister  rust  infestation  at  year  100  with  150-year 
stochastic  interval  fires. 

Fire-free  intervals  of  150  years  approximate  historical  fire 
frequency  in  most  BMWC  whitebark  pine  stands  (Keane 
and  others,  in  preparation).  This  stochastic  interval  im- 
plies there  is  a  0.0067  probability  of  a  wildfire  occmring 
in  any  simiilation  year.  Predictions  of  species  basal  area, 
fire  behavior,  and  fuel  loadings  were  output  from  the 
model. 

FIRESUM  was  then  tested  using  stand  stmctxiral  data 
from  adjacent  postfire  and  mature  stands  for  the  Cliff 
Moimtain  and  Tilison  Mountain  sampHng  sites.  TiHson 
Mountain  is  located  in  the  south-central  portion  of  the 
BMWC  near  the  Continental  Divide.  Both  sites  were 
bumed  in  the  1910  fire  that  bumed  much  of  the  BMWC. 
However,  portions  of  the  sites  escaped  the  fire  and  stand 
data  from  these  areas  were  used  as  inputs  to  the  model. 
Observed  conditions  in  the  post- 19 10  stands  were 
compared  with  FIRESUM  predictions  after  80  years  of 
simulation. 


247 


Table  1 — Average  characteristics  of  whitebark  pine  stands  types  in  the  Bob  Marshall  Wilderness 
Complex 


Average  Average 

percent  Average               percent  Average 

Tree                     overstory  overstory  understory  understory 

species                  basal  area'            age^  basal  area  age 


Whitebark  pine  50  258  21  96 

Subalpinefir  21  161  67  116 

Engelmann  spruce  16  228  6  101 


'Average  percent  of  basal  area  by  species  in  plots  within  each  site  type. 
'Average  age  of  each  species  in  plots  across  each  site  type. 


FIELD  STUDY  RESULTS 

Summarized  data  from  106  sample  sites  (fig.  1)  show 
whitebark  pine  stands  consist  of  an  overstory  of  old  white- 
bark pine  and  yoimger  subalpine  fir  and  spruce  with  an 
imderstory  of  mostly  subalpine  fir  and  spruce  (table  1). 
This  is  consistent  with  descriptions  by  Arno  (1986)  and 
Kendall  and  Amo  (1990)  to  the  effect  that  whitebark  pine 
is  being  successionally  replaced  by  the  more  shade- 
tolerant  fir  and  spruce.  Downed  woody  fuels  are  scant 
(less  than  5.5  kg  /m^);  most  (approximately  5.0  kg/m^)  are 
downed  logs  that  decompose  very  slowly.  Approximately 
70  percent  of  down  dead  tree  biomass  is  apparently 
whitebark  pine. 

Very  little  evidence  of  extensive  mountain  pine  beetle 
epidemics  was  observed  in  the  study  area.  Beetles  seem 
to  play  the  role  of  secondary  colonizer,  infecting  already 
stressed  pines  and  ultimately  contributing  to  their  death. 

Evidence  of  blister  rust  was  present  in  all  but  three  of 
106  sample  stands.  Blister  rust  infestation  levels  aver- 
aged around  80  percent  with  an  average  of  10  to  15  cank- 
ers per  tree  and  33  percent  of  tree  foliage  killed  by  the 
rust.  Severity  of  blister  rust  infestations  was  related  to 
geographic  area  but  not  correlated  with  topography  or 
serai  stage  (table  2). 


Blister  rust  is  prevalent  over  the  entire  BMWC  with  the 
highest  incidences  observed  in  the  northern  and  western 
portions  of  the  study  area  (fig.  2a-c).  Whitebark  pine 
stands  with  the  least  number  of  cankers,  portions  of  crown 
killed,  and  percents  of  trees  infected  appear  to  be  in  the 
southern  end  of  the  BMWC  along  the  Continental  Divide. 
Rust  severity  increases  as  one  goes  north,  west,  and  east 
of  the  Continental  Divide. 

Fire  history  was  difficult  to  determine  in  the  BMWC  be- 
cause of  the  rarity  of  fire-scarred  trees  in  the  whitebark 
pine  zone.  It  appears  stand-replacement  wildfires  often 
bum  trees  that  contain  fire  scars  and  leave  few  fire  his- 
tory records  on  the  ground.  However,  an  approximate  fire 
history  was  determined  from  stand  structiire  and  the  few 
fire  scars  foimd.  The  fire-free  intervsd  across  all  sample 
sites  was  approximately  144  years  with  a  minimimi  of  55 
and  a  maximum  of  304  years  (Keane  and  others,  in  prepa- 
ration). Gabriel  (1976)  foimd  that  the  Danaher  drainage 
of  the  southern  BMWC  had  an  average  stand-replacement 
fire  rotation  of  150  to  200  years.  Stand  age  for  all  sample 
sites  averaged  approximately  250  years  with  less  than 
1  percent  of  these  sites  experiencing  a  fire  in  that  time 
period. 


Table  2— Bob  Marshall  Wilderness  Complex  blister  rust  severity  by  geographical  area.  General  geograph- 
ical boundaries  are  shown  in  figure  1  and  are  as  follows:  Swan  Front— areas  west  of  South  Fork 
of  Flathead  River,  Continental  Divide— areas  between  South  Fork  Flathead  River  and  North 
Fork  Sun  River  and  south  of  Middle  Fork  Flathead  River,  Sawtooth  Front— areas  east  of  North 
Fork  Sun  River,  Middle  Fork— areas  in  the  Middle  Fork  Flathead  River  drainage  (northern 
BMWC).  Blister  rust  severity  is  expressed  as  a  percent  of  total  trees  infected  with  rust,  average 
number  of  cankers  per  tree,  and  percent  of  crown  killed  by  rust 


Geographical 
area 

Number 
plots 

Percent 
trees  infected 

Average 
cankers/tree 

Percent 
crown  kill 

Swan  Front 

38 

92 

14 

41' 

Continental  Divide 

33 

'67* 

9* 

15* 

Sawtooth  Front 

17 

86* 

13 

33* 

Middle  Fork 

18 

93 

14 

48* 

'*  =significant  at  p  =  0.05  using  Kruskal-Wallis  test  within  a  rust  severity  measure. 


248 


Wast 
Glacier 


i  KALISPELL 


80% 


N 
\ 
/ 

../...- 


70%  / 

/ 
/ 

;■•••/••■••■ 


50% 

/  I 

■y   ■■■  I 


.     ;   ,    \  ;  I 


50% 

! 


10%  V 


3%l  :  V 
/  : 


/  \ 

I  10%  *• 
\  • 


I 


i 


V. 
\V. 
Seeley  ^ 
Lake  0  y 


••:\  I 

50% 


Choteau 


•  Augusta 


West 

Glacier 


•  KALISPELL 


100%  \ 


/         >%  /  : 


:  80%  / 


100% 


100%       \  I  i\\ 

Seeiey  **•••*'».... 
Lake  •  \  ••»,. 


Choteau 


•  Augusta 


Figure  2 — Isometric  maps  showing  various  levels  of  three 
measures  of  blister  rust  severity  across  the  study  area:  a, 
blister  rust  infection  levels  (percent  of  trees  infected  by  blis- 
ter rust);  b,  blister  rust  canker  levels  (1 : 1-5,  2:  5-10,  3: 10- 
15,  4: 16-20,  5:  21-25  cankers  per  tree);  c,  percent  foliage 
killed  by  blister  rust  (estimated  portion  of  live  crown  lost  to 
blister  rust). 


249 


SIMULATION  STUDY  RESULTS 

FIRESUM  testing  results  showed  that  the  model  per- 
formed moderately  well  in  predicting  successional  stand 
composition.  Predicted  basal  areas  for  whitebark  pine 
and  subalpine  fir  sifter  80  years  of  simulation  were  within 
about  20  to  30  percent  of  those  observed  in  the  actual 
post-1910  fire  stands. 

Whitebark  pine  stand  dynamics  differed  greatly  across 
the  four  modeling  scenarios.  Predicted  whitebark  pine 
(PIAL)  basal  area  remained  somewhat  constant  under  a 
stochastic,  150-year  historical  fire  regime  with  subalpine 
fir  (ABLA)  and  Engelmann  spruce  (PIEN)  present,  but  at 
lower  levels  (fig.  3a).  However,  whitebark  pine  basal  area 
tended  to  decrease  somewhat  in  the  absence  of  fires  while 
subsdpine  fir  doubled  its  basal  area  (fig.  3b).  Whitebark 
pine  decline  is  greatly  accelerated  by  blister  rust  infesta- 
tions in  the  absence  of  fires  (fig.  3c).  Last,  fire  does  not 
seem  to  affect  the  decrease  in  whitebark  pine  levels  after 
blister  rust  introduction  (fig.  3d). 

DISCUSSION 

Historical  BMWC  upper  subalpine  stands  were  most 
likely  dominated  by  whitebark  pine  with  a  small  compo- 
nent of  fir  and  spruce.  Presence  of  an  older  age  class  of 
whitebark  pine  (table  1),  and  the  preponderance  of 
whitebark  pine  snags  (table  3)  and  downed  logs,  indicate 
that  these  stands  once  supported  about  20-30  m^/ha  of 
whitebark  pine  and  very  little  spruce  and  fir. 


At  present,  whitebark  pine  is  rapidly  declining  through- 
out most  of  the  BMWC.  Blister  rust  is  killing  many 
whitebark  pine  trees  along  the  western,  eastern,  and 
northern  boundaries  (fig.  2a-c).  A  1991  remeasurement 
of  whitebark  pine  trees  in  1971  vegetation  classification 
plots  indicated  that  approximately  20  percent  of  the  spe- 
cies' basal  area  and  30  percent  of  its  trees  are  lost  each 
decade  due  to  blister  rust  (Keane  and  Arno  1993).  Ab- 
sence of  whitebark  regeneration  in  sample  macroplots 
(table  1)  indicated  fir  and  spruce  will  eventually  replace 
whitebark  pine  without  fire. 

Lack  of  fire  on  the  landscape  has  allowed  subalpine  fir 
and  Engelmann  spruce  to  dominate  the  understory  of 
forests  previously  dominated  mostly  by  whitebark  pine 
(table  1).  As  a  result  of  fire  suppression,  few  subalpine 
areas  have  been  opened  by  fire  for  nutcracker  dispersal 


Table  3 — Standing  live  and  dead  tree  densities  (trees/ha)  by 

species  for  whitebark  pine  stands  in  the  Bob  Marshall 
Wilderness  Complex 

Tree  Live  tree   Live  tree   Dead  tree  Dead  tree 

species  density   basal  area    density    basal  area 


Trees/ha 

Whitebark  pine  21 8 
Subalpine  fir  490 
Engelmann  spruce  61 


m^/ha  Trees/ha  irf/ha 

15.0            43  5.2 

5.3            23  3.7 

5.7             2  .4 


250 


50 
40 


30 

(0 

o 

<  20 


(0 

(0  10 
CD 


a 


50  r 


I  40 

CM 

E 

f  20 

(S  10 


150-year  Fire  Interval 


RIAL 

ABLA 

PIEN 

\.    ■•    .-  ■•  •• ■■ 

1 

100      200  300 
Years 

Blister  Rust 


400  500 


RIAL 

ABLA 

RIEN 


100 


200  300 
Years 


400  500 


No  Fire 


RIAL 
ABLA 
RIEN 


100      200      300     400  500 
Years 


Blister  Rust/Fire 


50  r 
40 

CM 


ca 

0) 


30 
<  20 

(0 

S  10 


RIAL 

ABLA. 

RIEN 


100      200  300 
Years 


400  500 


Figure  3 — FIRESUM  simulation  results  for  Cliff  Mountain  site  under  the  four  modeling  scenarios:  a,  150- 
year  stochastic  fire  regime;  b,  no  fires  (fire  suppression);  c,  blister  rust  infection  at  year  100;  d,  150-year 
stochastic  fire  regime  and  blister  rust  infection  at  year  100. 


of  whitebark  pine  seeds.  Most  sampled  stands  were  older 
(>250  years)  than  the  estimated  fire  return  interval  (144 
years)  indicating  that  a  majority  of  whitebark  pine  stands 
have  exceeded  the  expected  fire  rotation. 

Fires  were  an  important  disturbance  on  the  BMWC 
landscape.  Extensive,  stand-replacement  fires  such  as 
those  documented  by  Ayers  (1901)  created  burned  areas 
colonized  by  whitebark  pine  seedlings  from  nutcracker- 
cached  seed.  These  large  fires  often  occurred  in  heavy 
fuels  during  extreme  dry,  windy  weather  conditions 
(Losensky  1990).  Fires  started  in  subalpine  areas  dvuing 
moderate  weather  years  consumed  scattered  surface  fuels 
and  often  did  not  ignite  tree  crowns.  However,  these  sur- 
face fires  generally  killed  fire-intolerant  fir,  spruce,  and 
young  whitebark,  but  only  scarred  the  older,  large 


whitebark  pine  trees.  This  type  of  fire  regime  created 
open,  parklike  stands  of  nearly  pure  whitebark  pine  (Amo 
1986). 

Model  results  show  that  future  BMWC  subalpine  for- 
ests will  probably  be  composed  mostly  of  fir  and  spruce 
with  Httle  whitebark  pine  (fig.  3c-d).  Introduction  of  fire 
into  the  BMWC  may  not  accelerate  the  loss  of  the  species 
at  the  stand  level,  but  may  help  perpetuate  the  presence 
of  whitebark  pine  on  the  BMWC  landscape  level  because 
burned-over  areas  are  better  nutcracker  caching  sites. 
Data  summaries  of  whitebark  pine  classification  plot 
remeasurements  (Keane  and  Amo  1993)  agree  with 
FIRESUM  modeling  results  that  predict  severe  reduction 
in  whitebark  pine  densities  40-50  years  after  blister  rust 
introduction  (fig.  3c). 


251 


CONCLUSIONS 

Whitebark  pine  popiilations  in  the  BMWC  are  decreas- 
ing at  an  alarming  rate.  Most  stands  will  probably  be 
converted  to  dense  subalpine  fir  and  spruce  with  minor 
components  of  whitebark  pine.  However,  cold,  dry  sites 
where  whitebark  pine  is  the  indicated  climax  may  be  con- 
verted to  shrub  or  herbaceous  commvmities;  this  may  ad- 
versely affect  snow  retention  and  watershed  dynamics 
(Keane  and  others  1990b).  Areas  currently  with  low  lev- 
els of  blister  rust  (such  as  the  southern  end  of  the  Conti- 
nental Divide)  will  probably  also  experience  heavy  white- 
bark pine  mortality,  though  much  more  slowly  than  other 
regions  due  to  more  infrequent  weather  conditions  con- 
ducive to  bUster  rust  infection  and  growth  (Hagle  and 
others  1989). 

Results  of  this  study  could  be  extrapolated  to  other 
parts  of  whitebark  pine's  range  that  experience  similar 
weather  patterns.  Whitebark  pine  population  levels  in 
Glacier  National  Park,  USA,  have  decreased  drastically 
in  the  last  20  years  (Kendall  and  Arno  1980)  due  mainly 
to  blister  rust.  Amo  (1986)  observed  a  decline  in  white- 
hark  pine  in  the  Bitterroot  Range  of  Montana  and  Idaho, 
USA.  Blister  rust  has  been  docimaented  in  the  southern 
portions  of  whitebark  pine's  range,  but  the  mortality  has 
not  been  as  extensive,  presumably  due  to  drier  weather 
conditions  (Carson  1978).  However,  given  suitable 
weather  conditions,  blister  rust  may  infect  many  more 
trees  in  this  drier  portion  of  its  range. 

Reduction  of  whitebark  pine  cone  crops  could  affect 
many  species  of  wildlife.  Grizzly  and  black  bears  must 
either  migrate  or  find  a  new  source  of  prehibernation 
foodstuffs  (Craighead  and  others  1982;  Kendall  1980; 
Mattson  and  Reinhart  1986).  Squirrels  and  Clark's  nut- 
crackers will  also  need  to  find  alternate  food  som-ces,  as 
will  other  animals  dependent  on  them.  Shifts  in  vegeta- 
tion composition  and  wildlife  migration  can  cause  major 
changes  in  landscape  diversity  and  structure.  In  turn, 
this  may  affect  the  pattern  of  fire  processes. 

About  1  to  8  percent  of  whitebark  pine  populations  may 
be  genetically  resistant  to  blister  rust  (Amo  and  Hoff 
1989;  Bingham  1972;  Hoff  and  others  1980).  This  might 
allow  whitebark  pine  to  remain  on  the  landscape  at  very 
low  levels.  However,  the  combined  effects  of  blister  rust 
and  successional  replacement  due  to  fire  suppression  will 
make  it  impossible  to  maintain  current  whitebark  pine 
population  levels  even  with  high  rust  resistance.  Also,  if 
these  populations  become  very  small,  nutcracker  seed 
consiunption  during  late  summer  could  destroy  most  of 
the  seed  crop  (Tomback  1982).  Breeding  rust-resistant 
populations  will  be  important  for  maintaining  whitebark 
pine  in  the  critical  portions  of  its  range,  especially  where 
it  is  currently  a  major  food  for  the  grizzly  bear.  Encour- 
aging nutcracker  caching  by  opening  dense  stands  with 
fire  may  also  promote  whitebark  pine  populations  (Sund 
and  others  1991). 

ACKNOWLEDGMENTS 

The  authors  thank  personnel  of  the  Spotted  Bear  Dis- 
trict of  the  Flathead  National  Forest  and  the  Rocky 


Moxmtain  District  of  the  Lewis  and  Clark  National  Forest 
for  all  the  aid  and  assistance  offered  throughout  the 
project.  Thanks  to  James  Menakis  and  Steve  Arno  of  the 
Intermountain  Research  Station  for  their  imlimited  help; 
and  Wendel  Hann  of  Forest  Service,  Northern  Region, 
Ecology  for  his  never-ending  support. 

REFERENCES 

Amo,  S.  F.  1970.  Ecology  of  alpine  larch  (Larix  lyallii) 
in  the  Pacific  Northwest.  Missoula,  MT:  University  of 
Montana.  264  p.  Dissertation. 

Amo,  S.  F.  1986.  Whitebark  pine  cone  crops:  a  diminish- 
ing source  of  wildlife  food?  Westem  Joumsd  of  Applied 
Forestry.  1(3):  92-94. 

Amo,  S.  F.;  Hoff,  R.  J.  1989.  Silvics  of  whitebark  pine  (Pi- 
nus  albicaulis).  Gen.  Tech.  Rep.  INT-253.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station.  11  p. 

Amo,  S.  F.;  Sneck,  K  M.  1977.  A  method  for  determining 
fire  history  in  coniferous  forests  of  the  Moimtain  West. 
Gen.  Tech.  Rep.  INT-42.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermoimtain  Forest 
and  Range  Experiment  Station.  28  p. 

Ayres,  H.  B.  1901.  Lewis  and  Clark  forest  reserve, 
Montana.  21st  Annual  Report.  Part  5.  Washington,  DC: 
U.S.  Department  of  the  Interior,  Geological  Siirvey: 
27-80. 

Bain,  S.  1989.  LANDSAT  classification  and  field  correla- 
tion methods  for  vegetation  classification  used  at  the 
Flathead  National  Forest.  In:  Ferguson,  D.  E.;  Morgan, 
P.;  Johnson,  F.  D.,  comps.  Proceedings — land  classifica- 
tion based  on  vegetation:  applications  for  resource  man- 
agement; 1987  November  17-19;  Moscow,  ID.  Gen.  Tech. 
Rep.  INT-257.  Ogden,  UT:  U.S.  Department  of  Agricul- 
tiu*e.  Forest  Service,  Intermountain  Research  Station: 
255-258. 

Bedwell,  J.  L.;  Childs,  T.  W.  1943.  Susceptibility  of  white- 
bark pine  to  blister  mst  in  the  Pacific  Northwest.  Jour- 
nal of  Forestry.  41:  904-912. 

Bingham,  R.  T.  1972.  Taxonomy,  crossability,  and  relative 
blister  mst  resistance  of  5-needled  white  pines.  In:  Biol- 
ogy of  mst  resistance  in  forest  trees.  Misc.  Publ.  1221. 
Washington,  DC:  U.S.  Department  of  Agriculture,  For- 
est Service:  271-280. 

Brown,  James  K  1974.  Handbook  for  inventorying  downed 
woody  fuel.  Gen.  Tech  Rep.  INT- 16.  Ogden,  UT:  U.S. 
Department  of  Agriculture,  Forest  Service,  Intermoim- 
tain Forest  and  Range  Experiment  Station.  24  p. 

Carlson,  C.  E.  1978.  Noneffectiveness  of  Ribes  eradication 
as  a  control  of  white  pine  blister  mst  in  Yellowstone 
National  Park.  Forest  Insect  and  Disease  Mgmt.  Rep. 
78-18.  Missoula,  MT:  U.S.  Department  of  Agricultiu-e, 
Forest  Service,  Northem  Region.  6  p. 

Ciesla,  W.  M.;  Furniss,  M.  M.  1986.  Idaho's  haunted  for- 
ests. American  Forests.  81(8):  32-35. 

Craighead,  J.  J.;  Summer,  J.  S.;  Scaggs,  G.  B.  1982.  A  de- 
finitive system  for  analysis  of  grizzly  bear  habitat  and 
other  wilderness  resources  utilizing  LANDSAT  multi- 
spectral  imagery  and  computer  technology.  Wildlife- 
Wildlands  Inst.  Monogr.  1.  Missoula,  MT:  University  of 
Montana  Foundation,  University  of  Montana.  279  p. 


252 


Femer,  J.  W.  1974.  Habitat  relationships  of  Tamiasciurus 
hudsonicus  and  Sciurus  aberti  in  the  Rocky  Mountains. 
Southwest  Naturalist.  18:  470-473. 

Gabriel,  H.  W.  1976.  Wilderness  ecology:  the  Danaher 
Creek  drainage,  Bob  Marshall  Wilderness,  Montana. 
Missoula,  MT:  University  of  Montana.  224  p.  Thesis. 

Hagle,  S.  K;  McDonald,  G.  I.;  Norby,  E.  A.  1989.  White 
pine  blister  rust  in  northern  Idaho  and  western 
Montana:  alternatives  for  integrated  management.  Gen. 
Tech.  Rep.  INT-261.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermovmtain  Research 
Station.  35  p. 

Hann,  W.  J.  1990.  Landscape  ecology  and  vegetation  man- 
agement in  whitebark  pine  ecosystems.  In:  Proceedings 
of  the  sympositmi — whitebark  pine  ecosystems:  ecology 
and  management  of  a  high  moxmtain  resource;  1989 
March  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  335-340. 

Hann,  W.  J.;  Jensen,  M.  E.;  Keane,  R.  E.  1988.  Chapter  4: 
Ecosystem  management  handbook — ECODATA  meth- 
ods and  field  forms.  Unpubhshed  paper  on  file  at:  U.S. 
Department  of  Agriculture,  Forest  Service,  Northern 
Region,  Missoula,  MT.  75  p. 

Hoff,  R.;  Bingham,  R.;  McDonald,  G.  J.  1980.  Relative 
bhster  rust  resistance  of  white  pines.  European  Journal 
of  Forest  Pathology.  10:  307-316. 

Hoff,  R.  J.;  McDonald,  G.  I.  1977.  Different  susceptibihty 
of  19  white  pine  species  to  woolly  aphid  (Pineus  colo- 
radensis).  Res.  Note  INT-225.  Ogden,  UT:  U.S.  Depart- 
ment of  Agricultiore,  Forest  Service,  Intermountain  For- 
est and  Range  Experiment  Station.  4  p. 

Hum,  J.  1989.  GPS— a  guide  to  the  next  utility. 
Siinnjrvale,  CA:  Trimble  Navigation.  76  p. 

Keane,  R.  E.;  Arno,  S.  F.  1993.  Rapid  decline  of  whitebark 
pine  in  western  Montana:  evidence  from  20-year  meas- 
xirements.  Western  Journal  of  AppUed  Forestry.  [In 
press]. 

Keane,  R.  E.;  Arno,  S.  F.;  Brown,  J.  K.  1989.  FIRESUM— 
an  ecological  process  model  for  fire  succession  in  west- 
em  conifer  forests.  Gen.  Tech.  Rep.  INT-266.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermountain  Research  Station.  76  p. 

Keane,  R.  E.;  Amo,  S.  F.;  Brown,  J.  K.;  Tomback,  D.  F. 
1990a.  Modelling  stand  dynamics  in  whitebark  pine  for- 
ests. Ecological  Modelling.  51:  73-95. 

Keane,  R.  E.;  Arao,  S.  F.;  Brown,  J.  K.  1990b.  Modeling 
distiirbance  and  conifer  succession  in  whitebark  pine 
forest.  In:  Proceedings  of  the  sjmaposium — whitebark 
pine  ecosystems:  ecology  and  management  of  a  high 
movuitain  resource;  1989  March  29-31;  Bozeman,  MT. 


Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  274-289. 
Keane,  R.  E.;  Jensen,  M.  E.;  Harm,  W.  J.  1990c. 
ECODATA  and  ECOPAC— analytical  tools  for  inte- 
grated resource  management.  The  Compiler.  8(3): 
24-37. 

Keane,  R.  E.;  Morgan,  P.;  Menakis,  J.  [In  preparation]. 
Status  of  whitebark  pine  {Pinus  albicaulis)  on  the  Bob 
Marshall  Wildemess  Complex  landscape. 

Kendall,  K.  C.  1980.  Bear-squirrel-pine  nut  interaction. 
In:  Yellowstone  grizzly  bear  investigations.  Annual  Re- 
port 1978-1979.  West  Glacier,  MT:  U.S.  Department  of 
the  Interior,  National  Park  Service:  51-60. 

Kendall,  K.  C;  Arno,  S.  F.  1990.  Whitebark  pine— an  im- 
portant but  dangered  wildlife  resovu-ce.  In:  Proceed- 
ings— symposium  on  whitebark  pine  ecosystems:  ecol- 
ogy and  management  of  a  high-moimtain  resource;  1989 
March  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agric\ilt\ire,  Forest  Ser- 
vice, Intermountain  Research  Station:  264-274. 

Losensky,  J.  1990.  A  comparison  of  the  1988  fire  season 
to  the  historical  role  of  fire  for  the  Bob  Marshall-Great 
Bear-Scapegoat  wildemess  complex.  Unpublished  re- 
port on  file  at:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermoimtain  Research  Station,  Intermoun- 
tain Fire  Sciences  Laboratory,  Missoula,  MT.  22  p. 

Mattson,  D.  J.;  Reinhart,  D.  P.  1986.  Grizzly  bear,  red 
squirrels,  and  whitebark  pine:  second  year  progress 
report.  Report  submitted  to  Interagency  Grizzly  Bear 
Study  Team.  Bozeman,  MT:  U.S.  Fish  and  Wildlife  Ser- 
vice. 38  p. 

Moore,  W.  R.  1984.  Last  of  the  Bitterroot  grizzly.  Montana 
Magazine.  68:  8-12. 

Sund,  S.  K;  Tomback,  D.  F.;  Hoffman,  L.  1991.  Post-fire 
regeneration  of  Pinus  albicaulis  in  western  Montana: 
pattems  of  occurrence  and  site  charactertistics.  Unpub- 
lished report  on  file  at:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Research  Station,  Inter- 
mountain Fire  Sciences  Laboratory,  Missoula,  MT.  55  p. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.;  Hoffman,  L.  A.;  Sund,  S.  K  1990.  Coevo- 
lution  of  whitebark  pine  and  nutcrackers:  implications 
for  forest  regeneration.  In:  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-moimtain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermoimtain  Re- 
search Station:  118-130. 


253 


SOME  ASPECTS  OF  CEMBRAN  PINE 
REGENERATION  IN  THE  ITALIAN 
COTTIANALPS 

Renzo  Motta 
Alberto  Dotta 


Abstract — In  the  Susa,  Chisone,  and  Varaita  Valleys  a  sponta- 
neous diffusion  of  cembran  pine  {Pinus  cembra)  has  been  ob- 
served during  the  last  few  decades  in  various  t3T)es  of  environ- 
ments and  at  altitudes  between  1,100  and  2,850  m.  The  initial 
height  growth  of  this  species  is  very  slow.  A  mean  age  of 
44  years  at  a  height  of  120  cm  was  found  in  86  cembran  pines 
taken  from  10  different  sites.  Regeneration  does  not  show  seri- 
ous problems  that  could  be  attributed  to  the  influences  of  fungi 
and  insects  or  climatic  factors.  The  main  cause  of  death  is  the 
damage  caused  by  game,  in  particular  fraying  damage  caused  by 
red  deer  (Capreolus  capreolus)  and  roe  deer  iCervus  elaphus). 


In  the  western  Italian  Alps  the  cembran  pine  (Pinus 
cembra)  covers  a  limited  area  compared  to  its  potential 
natxiral  range.  The  reasons  behind  this  are  well  known 
(Filipello  and  others  1976;  Stern  1988)  and  may  be  simi- 
marized  as  follows: 

•  replacement  of  woodland  by  grazing  and  timberline 
depression; 

•  replacement  of  cembran  pine  by  other  species,  in  par- 
ticular larch  (Larix  decidua),  which  is  more  suitable  for 
multiple-use  forestry — particularly  for  livestock  grazing; 

•  over-exploitation. 

In  the  wake  of  the  profound  socio-economic  change  that 
has  occurred  during  the  last  few  decades,  human  activity 
in  the  alpine  valleys  has  also  radically  changed.  Decline 
of  forest  use  and  grazing  has  allowed  a  widespread,  spon- 
taneous regeneration  of  cembran  pine. 

The  aim  of  this  study  is  to  describe  the  main  environ- 
ments in  which  this  widespread  regeneration  has  taken 
place,  to  observe  the  speed  wdth  which  this  process  has 
come  about  by  analyzing  the  early  height  growth  of  this 
pine  (up  to  the  height  of  120  cm),  to  determine  the  influ- 
ence of  the  direct  sunlight  on  the  increasing  regeneration, 
and  to  highlight  the  major  problems  with  regeneration 
during  this  initial  stage  of  secondary  succession. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Renzo  Motta,  Dipartimento  di  Agronomia  Selvicoltura  e  Gestione  del 


Territorio,  University  degli  Studi  di  Torino.  Via  L.  Da  Vinci  44,  Grugliasco 
(TO),  I  10095;  Alberto  Dotta,  Consorzio  Forestale  Alta  Valle  di  Susa.  P.zza 
Mistral  7,  Oulx  (TO),  I  10056 


CEMBRAN  PINE  DISTRIBUTION 

The  regional  distribution  of  cembran  pine  is,  in  broad 
perspectives,  quite  well  known  (Bono  and  Barbero  1971; 
Hoffman  1970).  The  southern  limit  of  the  species  in 
Piedmont,  and  indeed  over  the  entire  mountain  range, 
is  located  on  the  slopes  of  Mount  Mongioie  in  the  Tanaro 
Valley  at  an  sdtitude  of  1,800  m  above  sea  level  on  a 
southern  exposure.  Proceeding  northward,  cembrsm  pine 
occurs  sporadically  in  the  Pesio  Valley  and  more  fre- 
quently, although  generally  concentrated  in  rocky  sites,  in 
the  Gesso,  Stura,  and  Maira  Valleys.  Cembran  pine  is  well 
represented  in  the  Varaita,  Chisone,  and  Susa  Valleys. 

The  presence  of  cembran  pine  in  the  more  northerly  sec- 
tor of  the  region  is  limited.  In  the  Insubric  region  it  is  al- 
most absent  for  climatic  reasons,  in  the  Sesia  Valley  only 
a  small  spontaneous  population  is  found  in  the  Vogna  Val- 
ley (Bertolani-Marchetti  1961),  while  there  are  two  popu- 
lations in  the  Ossola  Valleys:  one  in  the  Anzasca  Valley 
and  another  one  in  the  Formazza  Valley  (Tiraboschi  1964). 
However,  in  this  latter  area  old  tree  stumps,  beams  used 
for  roofs,  and  trunks  found  in  the  moiintain  lakes  could 
lead  one  to  suppose  that  cembran  pine  was  present  in  the 
past  in  areas  where  it  is  no  longer  to  be  found  (Falcini 
1989). 

Altogether,  in  Piedmont  pure  cembran  pine  forest  occu- 
pies an  acreage  of  approximately  1,500  ha  while  mixed 
forest  amounts  to  over  3,500  ha  (Regione  Piemonte,  Ipla 
1985).  The  majority  of  this  forest  area  (over  60  percent) 
is,  however,  concentrated  in  three  valleys  of  the  Cottian 
Alps  that  have  been  the  subject  of  our  observations:  the 
Varaita,  Chisone,  and  Susa  Valleys  (fig.  1). 

STUDY  AREA 

Toward  the  end  of  the  Susa  and  Chisone  Valleys,  sev- 
eral kinds  of  geological  substratimi  can  be  observed:  Calc- 
schists  are  most  common  but  also  quartzite,  gneiss,  mica 
schist,  serpentine,  and  limestone  are  represented.  The 
Varaita  Valley  is  characterized  mainly  by  metamorphic 
rocks  (gneiss,  quartzite,  mica  schist),  Monviso  diabases, 
and  by  a  wide  belt  of  calcschists. 

Susa  and  Chisone  Valleys  are  dry  areas  and  belong  to 
the  most  continental  region  in  the  Alps  (Richard  and 
Tonnel  1987).  At  some  sites,  annual  precipitation  is  less 
than  700  mm.  Rainfall  regime  is  of  Piedmont  equinoxial 
type  with  maxima  in  spring  (the  principal)  and  in  au- 
tumn. Winter  is  the  driest  season.  In  the  Varaita  Valley 
precipitation  is  more  uniform  throughout  the  year,  and 
annual  precipitation  is  760  to  1,300  mm. 


254 


A  -  Upper  Susa  Valley 

B  -  Upper  Chisone  Valley 

C  -  Upper  Varalta  Valley 

W  Cembran  pine  distribution  in  the  Alps 

Figure  1 — Geographical  distribution  of  cembran  pine  in  the  Alps  and  localization  of  the  study  sites 
(Filipello  and  others  1976;  modified). 


In  the  inner  part  of  Susa  and  Chisone  Valleys,  the 
most-represented  forest  species  is  the  larch.  Among  the 
other  species,  mostly  in  the  Susa  Valley,  Scots  pine  {Pinus 
sylvestris),  silver  fir  (Abies  alba),  and  spruce  (Picea  abies) 
occur.  Cembran  pine  forms  a  few  small  pure  stands,  espe- 
cially close  to  the  upper  forest  limit,  which  in  Susa  Valley 
is  located  above  2,400  m.  In  these  two  valleys  there  are 
also,  in  some  special  edaphic  conditions,  Swiss  mountain 
pine  (Pinus  uncinata)  stands.  Larch  is  also  the  most  com- 
mon species  in  the  Varaita  Valley,  extending  down  to 
lower  elevations  where  it  replaces  beech  (Fagus  sylvatica) 
or  is  mixed  with  beech  copses  or  mixed  broad-leaved  for- 
ests. Cembran  pine  can  be  either  pure  stands  (southern 
aspects  as  Aleve  forest)  or  mixed  with  larch  stands  (cooler 
aspects). 


METHODS 

Sampling  on  sample  plots  (table  1)  at  regular  distances 
(1.5  km)  along  the  contour  lines  of  1,500,  1,700,  1,900, 
and  2,100  m  (according  to  "3P"  method)  allowed  us  to  de- 
scribe sites  and  cembran  pine  regeneration  in  the  areas 
examined  and  to  identify  the  environments  preferred  for 
regeneration. 

We  were  able  to  identify  106  areas  in  the  Susa  Valley, 
31  in  the  Chisone  Valley,  and  34  in  the  Varaita  Valley.  In 
each  of  these  areas  we  have  analyzed  the  main  character- 
istics of  the  site  (aspect,  gradient,  geology,  vegetation)  and 
of  the  forest  (species,  height,  health,  damages). 

Other  research  has  been  carried  out  at  tree  limit  and 
the  upper  regeneration  limit  of  the  cembran  pine  (Piussi 


255 


Table  1 — Description  of  sample  plots 


Valley 

Site 

Altitude 

Aspect 

Geology 

Vegetation 

Meters 

Varaita 

Alev§  1 

1,800 

S-W 

Ophlolite 

Cembran  pine  stand 

Varaita 

Alev6  1 

2.200 

S-W 

Ophlolite 

Cembran  pine  stand 

Chisone 

Souch.  1 

1,500 

N 

Calcschist 

Mountain  larch  stand 

Chisone 

Souch.  2 

1,900 

N 

Calcschist 

Subalpine  larch  stand 

Chisone 

Souch.  3 

2,250 

N 

Calcschist 

Subalpine  larch  stand 

Susa 

Enter 

1,800 

E 

Calcschist 

Subalpine  larch  stand 

OUod 

KA   1  lino 

vv 

Ocrperuirio 

ouQdipiiio  idrcn  sianu 

Susa 

P.  Bosco 

2,050 

N-W 

Calcschist 

Cembran  pine  stand 

Susa 

Ruine 

2,000 

N-W 

Calcschist 

Subalpine  larch  stand 

Susa 

V.  Gimont 

2,150 

W 

Moraine 

Subalpine  larch  stand 

and  Schneider  1985).  Additional  research  has  been  con- 
ducted below  1,500  m  in  the  montane  forests  to  identify  the 
lower  altitudinal  limit  of  distribution  and  regeneration. 

Eight  to  10  samples  of  cembran  pine  (height  at  least 
120  cm)  were  taken  at  10  sites  that  had  been  delineated 
by  altitude  and  vegetation. 

Trees  of  "average"  growth  form  were  chosen.  These 
were  growing  either  singly  or  in  small  clusters  of  two  to 
four  individuals  in  areas  where  regeneration  was  abim- 
dant.  For  the  small  clusters,  a  single  tree  per  group  was 
selected,  which  represented  average  habitat  features  and 
looked  relatively  undisturbed  by  the  other  cluster  trees  as 
far  as  the  amount  of  svuilight  received  is  concerned. 


These  sample  trees  were  cut  at  the  soil  surface  and 
then  sectioned  every  10  cm.  The  diameter  of  each  section 
was  measured  and  the  annual  rings  counted. 

RESULTS 

In  general,  cembran  pine  regeneration  is  increasing 
considerably  at  present  (fig.  2). 

In  broad  terms,  three  types  of  environment  can  be  dis- 
tinguished in  which  cembran  pine  regeneration  is  spread- 
ing at  present.  These  are  outlined  below  in  order  of 
importance. 


Trees  (d.b.h.  >  12.5  cm)  Regeneration  (30  cm<  h  <  150  cm) 


]  Picea  abies         \       I  Pinus  sylv.  &  una.  Broadleaves  B 


Figure  2— Percentage  composition  of  trees  (d.b.h.  >  12.5  cm)  and  of  regeneration  (30  cm  <  h  <  150  cm) 
at  different  altitudes  in  the  Upper  Susa  Valley. 


256 


WcK)dland  and  Abandoned  Pasture  Land 


Table  2— Upper  altitude  limits  attained  by  cembra  pine  regeneration 


Cembran  pine  is  gradually  invading  areas  again  that 
were  cleared  by  humans  in  the  past.  Cembran  pine  re- 
generation is  ver>'  common,  in  particular  under  larch 
cover  (Motta  and  Dotta  1992). 

This  phenomenon  may  also  be  favored  by  the  thick 
layer  of  herbaceous  vegetation  on  the  larch  forest  floor, 
which  prevents  the  seeds  of  other  species  from  reaching 
the  mineral  soil.  Seeds  of  cembran  pine,  thanks  to  the  ac- 
tion of  the  nutcracker  {Nucifraga  caryvcatactes),  manage 
to  overcome  this  obstacle.  Furthermore,  in  these  popula- 
tions, seeds  of  cembran  pine  are  stronger  than  seeds  of 
other  conifers  because  of  comparatively  high  nutrient  re- 
serves. Therefore,  seedlings  of  cembran  pine  can  more 
easily  compete  with  the  herbaceous  vegetation  on  the  for- 
est floor  (Trepp  1981). 

This  process  is  generally  more  apparent  on  northern  as- 
pects, while  on  the  southern  aspects  there  are  big  differ- 
ences between  the  three  valleys. 

In  the  Varaita  Valley,  the  cembran  pine  forest  of  Aleve 
(from  "elvni,"  cembran  pine  )  occurs  primarily  on  the  south- 
em  exposure.  In  this  valley  cembran  pine  also  is  regener- 
ating under  the  cover  of  larch  stands,  both  on  southern 
and  on  northern  slopes. 

This  situation  is  in  sharp  contrast  to  those  in  the  Susa 
and  Chisone  Valleys,  w^here  regeneration  of  cembran  pine 
is  predominant  on  northern  slopes.  The  presence  of 
cembran  pine  on  the  southern  slope  of  the  Chisone  Valley 
of  the  hamlet  "Alleve"  within  the  municipality  of 
Pragelato  (approximately  1,800  m  above  sea  level  with 
an  exposure  very  similar  to  that  of  Aleve  in  the  Varaita 
Valley)  seems  to  point  to  the  fact  that  present  absence  of 
the  cembran  pine  on  the  southern  slopes  of  the  upper 
Chisone  Valley  is  likely  due  to  the  action  of  humans. 

In  the  Susa  Valley  cembran  pine's  establishment  on 
southern  aspects  is  Hmited  by  ecological  factors  (Giordano 
and  others  1974)  and  also  by  geological  factors:  On  south- 
em  aspects  serpentine,  quartzite,  and  limestone  are  ver>' 
common.  These  substrata  are  mainly  colonized  by  Pinus 
uncinata. 

Generally,  southern  slopes  have  likely  been  used  for  ag- 
riculture and  grazing  more  intensely  and  for  longer  peri- 
ods of  time  than  northern  slopes.  Furthermore,  the  natu- 
ral recolonization  of  these  slopes  is  more  difficult  than  it 
is  on  the  northem  slopes.  Actually,  cembran  pine  forests 
found  on  the  southern  slopes  show  a  remarkable  stability 
when  conditions  are  suitable  for  regeneration  (Beguin  and 
Theurillat  1982),  but  as  soon  as  they  are  disturbed  they 
reveal  a  low  resilience,  and  spontaneous  recolonization  of 
the  original  sites  appears  to  be  rather  difficult. 

Forest  Limit  and  Ridges 

At  the  forest  Umit  cembran  pine  is  reinvading  ground 
where  it  had  been  excluded. 

Cembran  pine  regeneration  is  also  expanding  on  ridges 
and  in  isolated  "gruppe"  (clusters)  that  originated  from 
seed  caches  of  the  nutcracker  and  are  gro\'.'ing  well  above 
the  vegetation  limits.  These  constitute  particularly  favor- 
able microsites,  even  though  located  within  the  alpine 
zone  (Ozenda  1985).  In  the  Susa,  in  the  Chisone,  and 


Valley  Site  Tree  limit     Forest  limit  Aspect 


Meters 


Varaita 

Rocca  Jarea 

2,750 

2,300 

S-W 

Varaita 

Reisasso 

2,680 

2,380 

S 

Varaita 

Roccio  Russo 

2,640 

2,350 

s 

Varaita 

Losetta 

2,810 

2,240 

s 

Varaita 

Ire  Chiois 

2,800 

2,150 

s 

Cliisone 

C.  Chardonnet 

2,550 

2,250 

N-W 

Chisone 

A!  berg  i  an 

2,500 

2,290 

N-W 

Susa 

Dormilleuse 

2,620 

2,460 

E 

Susa 

M.  Gimont 

2,620 

2,400 

E 

in  the  Varaita  Valleys  the  forest  hmits  reach  altitudes 
(table  2)  among  the  highest  of  the  entire  Alps  (above 
2,400  m). 

Montane  Forests 

This  type  of  expansion  (.table  3  )  is  less  frequent  com- 
pared \Wth  the  two  described  earlier.  It  is  present  spo- 
radically, however,  in  all  three  of  the  valleys  examined  in 
this  study  where  regeneration  of  the  cembran  pine  occurs 
at  high  altitudes  and  in  uncommon  forest  t>T)es. 

This  phenomenon  is  by  no  means  easy  to  interpret  and 
cannot  be  considered  separately  from  seed  dispersal  by 
the  nutcracker  (Mattes  1982),  Certainly,  because  of  the 
thick  and  dense  herbaceous  cover,  cembran  pine  seeds 
have  distinct  advantages  over  those  of  other  conifers,  as 
was  mentioned  in  the  discussion  of  the  subalpine  zone. 

Early  Height  Growth 

A  sur\'ey  of  the  number  of  years  young  cembran  pines 
need  to  grow  to  120  cm  in  height  has  been  carried  out 
in  order  to  know  the  duration  of  this  first  colonization 
period. 

Analysis  of  the  early  height  growth  of  cembran  pines 
has  already  been  conducted  in  other  areas  of  the  Alps 
(Contini  and  Lavarelo  1982;  Oswald  1963;  Unterricher 
1986)  by  different  authors  using  various  methodologies. 
Some  investigations  carried  out  in  the  Upper  Susa  Valley 


Table  3 — Lower  altitude  limits  attained  by  cembra  pine  regeneration 

in  the  montane  zone 


Valley 

Site 

Vegetation 

Altitude  Aspect 

Meters 

Varaita 

Confine 

Mountain  larch  stand 

1,150 

N 

Varaita 

Torrette 

Beech  stand 

1,250 

N 

Chisone 

Souch.  basses 

Mountain  larch  stand 

1,450 

N-W 

Chisone 

Fraisse 

Mountain  larch  stand 

1,480 

N-W 

Susa 

Meana 

Mountain  larch  stand 

1,200 

N 

Susa 

Mian 

Beech-fir  stand 

1,250 

N 

Susa 

Gran  Bosco 

Fir  stand 

1.300 

N-E 

257 


Table  4 — Number  of  years  required  to  reach  120  cm  height 


Mean  age 

Number  of 

at  1 20  cm 

Aae 

Standard 

Site 

trees 

height 

range 

deviation 

a)  Aleve  1 

9 

43 

28-71 

15.1 

b)  Aleve  2 

8 

45 

38-51 

4.7 

c)  Souch.  1 

8 

52 

36-73 

11.2 

H\  Snurh  2 

59 

49-68 

6.7 

e)  Souch.  3 

8 

40 

33-47 

3.9 

f)  Enter 

8 

39 

35-44 

2.9 

g)  M.  Luna 

9 

55 

36-77 

10.0 

h)  V.  Gimont 

9 

37 

26-66 

11.2 

i)  Ruine 

9 

39 

26-49 

7.6 

1)  P.  Bosco 

10 

oo 

<;4-40 

O.O 

Varaita  V.  (a,b) 

17 

44 

28-71 

11.2 

Chisone  V.  (c-e) 

24 

50 

33-73 

11.2 

Susa  V.  (g-l) 

45 

40 

24-77 

11.2 

Total 

86 

44 

24-77 

12.2 

have,  however,  provided  data  significantly  different  from 
those  pubHshed  in  Hterature.  At  the  same  time,  the  pres- 
ence of  cembran  pine  regeneration  in  such  different  situa- 
tions as  far  as  altitude,  aspect,  and  type  of  vegetation  are 
concerned  has  led  us  to  proceed  further  with  this  line  of 
research  to  see  whether  there  are  significant  differences 
in  the  early  height  growth  of  the  cembran  pine  imder  dif- 
ferent ecological  conditions. 

From  data  collected,  early  growth  appears  to  be  signi- 
ficantly slower  than  observed  by  other  authors  (table  4). 

In  the  Varaita  Valley,  the  age  of  young  cembran  pine 
at  120  cm  height  is  quite  regular  over  the  whole  slope. 
However,  growth  rates  slightly  decrease  with  increasing 
altitude. 


The  situation  in  the  Chisone  Valley  is  the  most  interest- 
ing. Three  sites  were  chosen  of  the  same  slope,  aspect, 
geology,  and  vegetation  (larch  stands),  but  at  different 
altitudes.  The  site  at  the  highest  altitude  shows  the  best 
growth  rates.  Probably  the  competition  of  the  abundant 
herbaceous  vegetation  on  the  two  lower  sites  reduced  the 
growth  of  cembran  pines  in  the  early  decades  of  their  life. 

In  the  Susa  Valley,  cembran  pines  grow  more  quickly 
than  in  the  other  two  valleys,  with  the  exception  of  the 
area  of  Monti  della  Lima.  There,  the  geological  substra- 
tiun  (serpentine)  and  the  shallow  soil  impede  the  growth 
of  plants  throughout  their  life. 

Regeneration  Problems 

The  Susa,  Chisone,  and  Varaita  Valleys  are  character- 
ized by  moderate  winter  snowfalls  and  early  snow  melt 
in  spring.  Consequently,  problems  related  to  Phacidium 
infestans  and  Herpotrichia  nigra  are  negligible.  In  the 
Susa  Valley,  the  endemic  Peridermium  strohi  does  not 
have  any  effect  on  adult  plants,  but  may  cause  young 
trees  to  wilt  and  lead  to  higher  mortality  among  seed- 
lings. Also,  problems  caused  by  insects  (in  particular  Ips 
cembrae)  and  frost  damage  are  negligible.  In  the  Varaita 
Valley,  three  fires  have  destroyed  several  hectares  of 
cembran  pine  woodland  during  the  last  few  decades,  but 
these  areas  have  promptly  been  recolonized  by  cembran 
pine. 

The  biggest  problem  regarding  regeneration  of  the 
cembran  pine  is,  at  present,  damage  caused  by  wildlife 
(fig.  3).  This  damage  is  not  distributed  uniformly  over 
the  area.  It  is  serious  in  the  Susa  Valley  (28.9  percent 
of  cembran  pine  regeneration  damaged),  substantial  in 
the  Chisone  Valley  (11.8  percent),  and  negligible  in  the 
Varaita  Valley,  where  for  now  there  are  no  standing  pop- 
ulations of  red  deer  (Ceruus  elaphus)  and  roe  deer  (Capre- 
olus  capreolus).  In  addition,  the  various  types  of  damage 


1 .3%  Black  grouse 


0.6%  Mice  and  Hare 
1.9%  Wild  boar 


5.7%  Red  deer  bark  stripping 


33.1%  browsing 
(Chamois,  Roe  deer, 
Red  deer) 


57.1%  fraying 

(Roe  deer,  Red  deer) 


Figure  3 — Wildlife  damaging  cembran  pines  in  the  Upper  Susa 
Valley. 


258 


do  not  have  the  same  effect,  and  trees  are  killed  mainly 
by  fraying  damage  caused  by  deer.  Overall,  cembran  pine 
mortality  is  due  primarily  to  damage  caused  by  animal 
behavior  (fig.  4). 

In  the  areas  examined,  cembran  pine  is  affected  only  to 
a  limited  extent  by  browsing  because  the  ungulates  prefer 
other  more  palatable  species  at  lower  altitudes.  With  re- 
spect to  tree  height  and  animal  browsing,  cembran  pine  is 
protected  by  snow  in  the  winter  months.  On  the  other 
hand,  saplings  in  the  height  range  susceptible  to  fraying 
damage  are  particularly  affected.  A  recent  study  (Motta 
and  Quaglino  1989)  has  shown  a  significant  incidence  of 
fraying  damage  in  the  cembran  pine  compared  to  other 
species  in  the  Upper  Susa  Valley.  To  explain  this  situa- 
tion, we  can  make  a  number  of  h3^otheses:  a  greater  pres- 
ence of  the  cembran  pine  in  areas  where  this  kind  of  dam- 
age is  more  likely  to  occur;  the  fact  that  the  cembran  pine 
remains  for  a  considerable  length  of  time  (at  least  60-80 
years)  within  a  size  range  prone  to  fraying  damage;  the 
elasticity  of  the  young  trunks  to  the  rubbing  action  of  the 
antlers;  the  abundance  of  intensely  scented  resin. 

In  the  Varaita  Valley,  wild  boars  (Sus  scrofa),  which 
usually  are  favorable  to  forest  regeneration,  cause  the 
death  of  a  large  number  of  cembran  pine  seedlings  by 
"plowing  up"  the  soil. 

The  ratio  of  dead  to  damaged  trees  due  to  wild  ungu- 
lates is  limited,  amounting  to  25.7  percent  of  damaged 


cembran  pine  regeneration  in  the  Susa  Valley,  12.8  per- 
cent in  the  Chisone  Valley,  and  less  than  1  percent  in  the 
Varaita  Valley.  However,  locally,  for  example,  in  the  ar- 
eas of  red  deer  rut,  this  damage  can  be  serious  (over  65 
percent  of  trees  damaged  and  a  25  percent  death),  cover- 
ing several  hectares. 

DISCUSSION 

The  spontaneous  diffusion  of  cembran  pine  in  the  Alps 
of  Piedmont  is  a  phenomenon  related  to  modifications 
caused  by  humans  in  the  past  and  to  the  present  decrease 
of  human  impact. 

This  phenomenon  could  be  favored  by  two  other  factors: 

•  Climatic  change  during  the  last  few  decades 
(Graumlich  1991;  Hansen-Bristow  1986); 

•  Modifications  of  certain  parameters  (numbers,  den- 
sity, range,  etc.)  and  seed  dispersal  by  the  nutcracker  af- 
fecting this  diffusion  positively. 

The  regeneration  of  cembran  pine  constitutes  an  asset 
from  the  silvicultural  standpoint  in  that  it  contributes  to 
the  renewal  of  the  topsoils  that  were  degraded  due  to  in- 
tense forest  use  by  humans.  For  example,  larch  stands 
used  for  grazing  are  not  likely  to  experience  natural 
regeneration,  and  artificial  regeneration  would  be  some- 
what problematic  and  very  expensive.  This  process  thus 


50 


]  Red  deer  bark  str.     mmm  Wild  boar 


Figure  4 — Cembra  pine  mortality  caused  by  wildlife. 


259 


increases  the  physical  £ind  ecological  stability  of  the  moun- 
tain forests  and  enhances  their  protective  and  naturalistic 
value. 

However,  vegetation  succession  over  very  long  time  in- 
tervals must  be  taken  into  consideration.  This  is  clearly 
evidenced  by  the  very  slow  early  height  growth  of  young 
cembran  pine.  It  would  require  but  a  few  springs  with 
heavy  and  prolonged  snowfalls  scattered  over  a  range  of 
several  dozen  years  to  favor  the  spread  of  pests  such  as 
snow  fungi,  and  thus  jeopardize  the  dynamic  rehabilita- 
tion process  that  has  taken  decades  to  accomplish.  How- 
ever, such  abundant  and  prolonged  snowfalls  occur  very 
rarely  in  the  valleys  that  have  been  the  subject  of  o;ir 
study,  but  for  this  very  reason  they  constitute  an  even 
greater  hazard.  Thus  we  consider  it  indispensable  to  pro- 
ceed with  the  monitoring  of  these  topsoils  to  keep  the  cur- 
rent factors  of  mortality  below  a  tolerable  level  and  to  fa- 
vor, wherever  possible,  the  formation  of  topsoils  that  are 
not  uniform  either  structurally  or  compositionally. 

REFERENCES 

Beguin,  C;  Theurillat,  J.  P.  1982.  La  foret  termophile 

d'arolles.  Candollea.  37:  349-379. 
Bertolani  Marchetti,  D.  1961.  Ricerche  sulla  vegetazione 

in  Valsesia.  5)  Pinus  cembra  L.  in  Valsesia  (Vercelli  and 

Piemonte).  Nuovo  Giornale  Botanico  Italiano.  68(3-4): 

1-3. 

Bono,  G.;  Barbero,  M.  1971.  A  propos  des  cembraies  des 
Alpes  cotiennes  italiennes,  maritimes  et  ligures. 
Allionia.  17:  97-120. 

Contini,  L.;  Lavarelo,  Y.  1982.  Le  pin  cembro  (Pinus 
cembra  L.):  vegetation,  ecologie,  sylviculture  et  produc- 
tion. Paris:  INRA.  197  p. 

Falcini,  L.  1989.  I  boschi  delle  valli  ossolane.  Torino: 
Assessorato  Agricoltiira  e  foreste  Regione  Piemonte. 
40  p. 

Fihpello,  S.;  Sartori,  F.;  Vittadini,  M.  1976.  Le  associa- 
zioni  del  Cembro  nel  versante  meridionale  dell'arco 
alpino  (introduzione  e  caratteri  floristici).  Atti  1st.  Bot. 
Lab.  Critt.  Univ.  Pavia:  21-104. 

Giordano,  A.;  Mondino,  G.  P.;  Palenzona,  M.;  Rota,  L.; 
Salandin,  R.  1974.  Ecologia  ed  utilizzazioni  prevedibili 
della  valle  di  Susa.  Ann.  1st.  Sper.  Arezzo.  5:  82-196. 

Graumlich,  L.  J.  1991.  Subalpine  tree  growth,  climate  and 
increasing  CO^  :  an  assessment  of  recent  growth  trends. 
Ecology.  72(1):  1-11. 


Hansen-Bristow,  K.  1986.  Influence  of  increasing  eleva- 
tion on  growth  characteristics  at  timberline.  Canadian 
Journal  of  Botany.  64(11):  2517-2523. 

Hoffman,  A.  1970.  L'areale  italiano  del  pino  cembro. 
Webbia.  25(1):  199-218. 

Mattes,  H.  1982.  Die  Lebensgemeinschaft  von 
Tannenhaner  und  Arve.  Eidgenossische  Anstalt  fur  das 
forstliche  Versuchswesen.  241.  74  p. 

Motta,  R.;  Quaglino,  A.  1989.  Sui  danni  provocati  dalla 
fauna  selvatica  ai  popolamenti  forestali  in  Alta  Valle  di 
Susa.  L'ltalia  Forestale  e  Montana.  44(4):  241-260. 

Motta,  R.;  Dotta,  A.  1992.  Le  dynamisme  des  melezeins 
des  Hautes  Vallees  de  Suse  et  du  Chisone  (Piemont, 
Italie).  Chambery:  Icalpe  Integralp  Final  Report.  [In 
print]. 

Oswald,  H.  1963.  Verteilung  und  Zuwachs  der  Zirbe 
(Pinus  cembra  L.)  der  subalpinen  Stufe  an  einem  zen- 
tralapinen  Standort.  Mitt,  der  forst.  Bundesversuch- 
sanst.  Mariabrunn.  60:  437-500. 

Ozenda,  P.  1985.  La  vegetation  de  la  chane  alpine.  Paris: 
Masson  Ed.  331  p. 

Piussi,  P.;  Schneider,  A.  1985. 1  limiti  superiori  del  bosco  e 
degli  alberi  in  val  di  Vizze  (prov.  di  Bolzano).  Ann.  Acc. 
It.  Sc.  For.  34:  121-150. 

Regione  Piemonte;  IPLA.  1982. 1  boschi  e  la  carta 
forestale  del  Piemonte.  Napoli:  Guida  ed.  177  p. 

Richard,  L.;  Tonnel,  A.  1987.  Contribution  a  I'etude  des 
vallees  internes  des  Alpes  Occidentales  (Premiere 
partie:  originalite  du  milieu  et  quelques  consequences 
biologiques).  Documents  de  Cartographie  Ecologique. 
30:  113-136. 

Stern,  R.  1988.  The  consequences  of  htmian  impact  on 
tree  borderlines.  In:  Saibitano,  F.,  ed.  Human  influ- 
ences on  forest  ecosystems  development  in  Europe: 
Symposivun  proceedings;  1988  September  26-29;  Trento. 
Bologna:  Pitagora  Ed.:  279-286. 

Tiraboschi,  G.  1964.  La  presenza  di  formazioni  spontanee 
di  Pino  cembro  e  Pino  uncinato  nelle  vallate  alpine 
novaresi  con  particolare  riferimento  alia  zona 
deirOssola.  Novara:  C.C.I.A.  di  Novara.  52  p. 

Trepp,  W.  1981.  Das  Besondere  des  Planterns  im 
Gebirgswald.  Schweiz.  Z.  Forstwes.  132(10):  823-846. 

Unterricher,  M.  1986.  II  cirmolo.  Note  ecologiche, 
vivaistiche  e  selvicolturali.  Milano:  Regione  Lombardia. 
114  p. 


260 


MIXED  CEMBRAN  PINE  STANDS 
ON  THE  SOUTHERN  SLOPE  OF 
THE  EASTERN  ALPS 

Pietro  Piussi 


Abstract — Cembran  pine  {Pinus  cembra)  dominates  over 
European  larch  {Larix  decidua)  and  Norway  spruce  (Picea  ahies) 
in  terms  of  species  composition,  basal  area,  and  regeneration. 
Spruce  and  larch  dominate  lower  and  medium  altitudes,  pine  the 
highest  elevations.  Due  to  intensive  grazing  and  logging  in  the 
past,  72  percent  of  trees  on  upper  timberline  are  under  100  years 
old;  only  0.25  percent  exceed  300  years.  Forest  limit  is  advanc- 
ing. The  state  of  health  is  good. 


Botanical  studies  have  been  xindertaken  in  woodlands 
near  the  upper  timberline  of  the  eastern  Alps  (Filippello 
and  others  1980),  but  very  little  is  known  about  the  struc- 
ture of  the  stands  (Del  Favero  and  others  1985;  Piussi  and 
Schneider  1985;  Unterrichter  1986). 

Traditionally,  the  upper  timberline  was  intensively 
grazed  during  the  summer,  a  practice  still  common  today 
although  numbers  of  domestic  animals  have  decreased. 
The  woods  were  logged  for  domestic  and  mining  purposes, 
trees  were  ring-barked,  and  often,  imtil  recent  times, 
seedhngs  were  uprooted  to  create  grazing  land.  Nowa- 
days, winter  sports,  as  well  as  high  deer  (Cervus  elaphus) 
and  chamois  (Rupicapra  rupicapra)  populations  in  the  ab- 
sence of  natural  predators,  exert  a  new  kind  of  pressure 
on  the  timberUne. 

Over  centuries,  these  activities  have  lowered  the  forest 
limit  and  substantially  modified  the  structure,  density, 
and  regeneration  patterns  of  high-altitude  stands. 

During  the  last  20  years,  the  Institute  of  Silviculture  of 
the  University  of  Florence  has  carried  out  a  series  of  stud- 
ies on  woodland  stands  near  the  upper  timberline  on  the 
southern  slope  of  the  eastern  Alps,  in  order  to  determine 
the  general  characteristics  of  conifer  stands  in  which 
cembran  pine  {Pinus  cembra)  is  present  and  sometimes 
dominant.  The  studies  established  the  position  of  forest 
limit  and  tree  line,  and  provided  data  on  woodland  stand 
structure  at  high  altitudes. 

This  paper  analyzes  the  results  of  10  of  these  studies 
in  an  attempt  to  reconstruct  past  dynamics,  determine 
present  conditions,  and  foresee  futiire  developments  in 
this  ecotone. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Pietro  Piussi  is  Professor  of  Silviculture,  Istituto  di  Selvicoltura, 
Universita'  degli  Studi  di  Firenze,  Via  S.  Bonaventura,  13,  50145 
FIRENZE,  Italy. 


SITE  DESCRIPTION 

Research  was  conducted  in  three  geographical  areas  in 
the  provinces  of  Trento,  Bolzano  and  Sondrio,  northern 
Italy,  (fig.  1): 

•  West:  the  Rhaetian  Alps  and  Valtellina  (valleys 
Viola,  Solda,  Furva,  and  Genova). 

•  Southeast:  the  Dolomites  and  central  areas  (valleys 
Fiemme,  Badia,  and  Fimes). 

•  Northeast:  near  the  Austrian  border  (valleys  Aurina 
and  Vizze,  Monte  Croce  Alta). 

The  climate  becomes  more  continental  going  fi'om  the 
south  and  west,  where  beech  (Fagus  sylvatica)  is  common 
at  lower  altitudes  and  silver  fir  (Abies  alba)  can  reach  the 
forest  limit,  to  the  north,  where  high-altitude  forests  are 
formed  by  Eiiropean  larch  (Larix  decidua),  cembran  pine 
and  Norway  spruce  (Picea  abies).  Precipitation  decreases 
from  south  to  north  and  its  regime  changes:  there  are 
spring  and  autumn  maxima  in  the  south  and  a  summer 
peak  in  the  north. 

The  terrain  is  rugged,  and  mountain  tops  surpass 
3,000  m  altitude. 


Figure  1 — Study  area  locations  a.-id  author 
references.  A,  Val  Viola  (Rapella  1985); 
B,  Val  di  Solda  (Travan  1984);  D,  Valfurva 
(Pirelli  1984);  E,  Val  Genova  (Zoanetti  1983); 
F,  Val  di  Fiemme  (Olivari  1983);  M,  Val  Badia 
(Catalano  1987);  N,  Val  di.Funes  (Maniero 
1987);  0,  Val  di  Vizze  (Piussi  and  Schneider 

1985)  ;  P,  Val  Aurina  (Hellweger  1989); 
Q,  Monte  Croce  Alta  (Piussi  and  Proietti 

1986)  . 


261 


METHODS 


Table  1 — Species  composition  on  sunny  and  shaded  slopes 


For  every  site,  three  determinations  were  made:  (1) 
forest  limit — the  upper  Hmit  of  woodland  above  500  m^ 
with  sufficient  density  to  create  a  certain  woodland  cli- 
mate inside  (Rubner,  in  Mayer  and  Ott  1991);  (2)  tree 
line  for  each  species — the  line  of  isolated  trees  above  2  m 
tall,  as  these  emerge  from  average  snow  cover  (Ellenberg, 
in  Mayer  and  Ott  1991);  and,  (3)  krummholz  limit — 
individuals  of  tree  species  above  the  forest  limit,  of  con- 
torted growth  and  below  2  m  tall  due  to  environmental 
factors  (Mayer  and  Ott  1991). 

Stand  conditions  were  examined  in  regularly  spaced 
500-m^  sample  plots  of  varying  aspect  and  altitude,  with 
the  highest  plots  some  30  to  100  m  below  the  forest  limit 
(average  1.5  ha  sampled  per  site;  total  sample  area  15.4  ha). 
In  each  plot,  some  recordings  were  done  for  all  individuals 
above  1.3  m  height  (species,  d.b.h.,  height  class,  health 
condition)  and  others  for  a  percentage  of  them  (height, 
age),  and  for  all  regeneration  of  0.2  to  1.3  m  height  (spe- 
cies, height  class,  health  condition). 

RESULTS 

As  the  research  evolved  over  a  period  of  2  decades,  per- 
spectives changed  and  some  surveys  carried  out  addi- 
tional work.  It  has  frequently  been  possible  to  separate 
portions  of  the  same  study  area  into  sunny  and  shaded 
aspects  and  into  different  altitudinal  bands.  Data  refer 
to  individuals  above  1.3  m  height  unless  otherwise  stated. 

Although  structure  was  analyzed  in  detail  at  a  local 
scale,  this  paper  describes  only  the  general  situation. 

Forest  Limit  and  Tree  Line 

Average  forest  limits  for  each  study  area  range  from 
2,080  to  2,325  m  above  sea  level  (a.s.l.)  (2,380  m  maxi- 
mvun),  with  considerable  local  variation  in  each  area. 
Cembran  pine  reaches  the  highest  elevations,  with  its 
tree  line  on  average  120  m  above  the  forest  limit  (larch 
tree  line  60  m,  and  spruce  tree  line  30  m  above  forest 
limit  or  absent). 

On  sites  with  different  aspects,  both  forest  limit  and 
cembran  pine  tree  line  are  generally  higher  on  sunny 
slopes,  by  20  to  30  m  and  about  45  m,  respectively.  These 
values  are  clearly  lower  than  those  given  by  Schroeter 
(1908)  and  Mayer  and  Ott  (1991),  who  emphasize  100  m 
and  more  difference  between  forest  limits  on  warm  and 
cold  slopes. 

Species  Composition 

Cembran  pine  forms  mixed  stands  with  spruce  and 
larch;  rowan  (Sorbus  aucuparia)  and  Scots  pine  (Pinus 
sylvestris)  are  sporadic.  The  behavior  of  larch  and  spruce 
will  be  considered  here  in  as  much  as  it  affects  the  propor- 
tion and  spatial  distribution  of  cembran  pine  (table  1). 
The  relative  percentage  of  each  species  depends  upon  site 
aspect  and  altitude. 

In  the  west,  cembran  pine,  dominant  nearly  every- 
where, prefers  warm  slopes  in  Solda/Furva,  whereas 


Sunny  slope   Shaded  slope 

Location    Pine     Spruce    Larch      Pine     Spruce  Larch 

 Percent  


V  lUla 

sn 

14. 

oOlUa 

1  O 

c 
o 

O  1 

o 

Furva 

87 

8 

5 

34 

66 

Geneva 

32 

68 

11 

89 

Fiemme 

47 

49 

4 

70 

27 

3 

Badia 

35 

18 

47 

Funes 

80 

15 

5 

71 

10 

19 

Vizze 

54 

14 

32 

Aurina 

83 

7 

10 

Croce 

15 

49 

36 

79 

21 

in  Viola  aspect  does  not  affect  species  composition.  Larch 
is  the  second  most  important  species. 

In  the  southeast,  in  Fiemme,  cembran  pine  is  codomi- 
nant  with  spruce  on  sunny  sites,  but  leads  on  shaded 
ones,  with  70  percent.  In  Fimes,  cembran  pine  dominance 
is  even  more  pronounced,  both  on  warm  and  cold  slopes. 
Again,  larch  increases  its  presence  on  cold  slopes,  while 
spruce  numbers  decrease.  By  contrast,  in  Badia,  nearly 
half  the  population  on  shaded  sites  consists  of  larch,  with 
pine  making  up  only  35  percent.  There  are  no  data  for 
sunny  slopes. 

In  the  northeast,  cembran  pine  dominates  the  south- 
exposed  slopes  of  Vizze  and  Aurina  (54  percent  and  83 
percent),  whereas  in  Croce  Alta  cembran  pine  constitutes 
15  percent  on  the  sunny  and  79  percent  on  the  shaded 
slopes. 

Average  values  can,  however,  obscure  significant  altitu- 
dinal variations,  as  is  the  case  for  Badia,  Aurina,  and 
Croce  Alta,  for  which  data  divided  into  distinct  altitudinal 
belts  are  available. 

In  Badia,  the  proportion  of  cembran  pine  increases 
slowly  from  30  percent  to  39  percent  at  altitudes  of  1,900 
to  2,100  m.  In  fact,  the  lower  belt  is  dominated  by  spruce, 
the  other  two  by  larch.  In  the  Aurina  plots,  along  four 
belts  at  1,870  to  2,340  m,  cembran  pine  numbers  increase 
irregularly  with  altitude:  from  71  percent,  through  86  per- 
cent and  76  percent,  to  100  percent.  Only  on  the  sunny 
side  of  Croce  Alta,  cembran  pine  decreases  considerably 
with  altitude:  from  22  percent  in  the  larch-spruce  domi- 
nated stand  at  1,900  m  to  9  percent  in  the  spruce  stand 
at  2,000  m.  In  the  higher,  shaded  belts,  cembran  pine 
dominates  spruce  at  2,100  m  and  forms  a  pure  stand  at 
2,200  m.  It  must  be  emphasized,  however,  that  those  on 
sunny  side  plots  are  distinctly  lower  than  those  on  shaded 
sides. 

Plant  Density  and  Basal  Area 

In  every  study  area,  total  plant  density  (number  of 
plants  per  hectare)  is  substantially  higher  on  sunny  than 
on  shaded  slopes  (table  2),  varying  between  800  and  900 
individuals  per  hectare  on  warm  and  500  and  600  individ- 
uals per  hectare  on  cold  sites.  However,  average  values 


262 


Table  2 — Plant  density  (number  of  plants  per  hectare)  and  basal  area  (m^  per  hectare) 


Sunny  slope    Shaded  slope 


Location 

Total 

Pine 

Spruce 

Larch 

Total 

Pinp 
■  II  ic 

Spruce 

1  Arch 

No. 

BA 

No. 

BA 

No. 

BA 

No. 

BA 

No 

BA 

No 

BA 

No.  BA 

No 

BA 

Viola 

909 

18 

455 

5 

130 

3 

324 

g 

685 

WW 

15 

113 

8 

P1Q 

1  P 

Solda 

792 

36 

651 

28 

99 

4 

42 

*tOv/ 

0\j 

9 

1 

P1  ^ 
c.  \  O 

Furva 

1,265 

30 

1 ,102 

26 

95 

2 

68 

2 

1  o 

Q 

o 

yjCO 

o 
o 

Genova 

454 

25 

144 

6 

310 

19 

386 

19 

40 

2 

346 

17 

Fiemme 

893 

60 

421 

32 

437 

24 

35 

4 

534 

38 

378 

29 

141 

8 

15 

1 

Badia 

439 

26 

153 

10 

82 

4 

204 

12 

Funes 

758 

29 

608 

21 

117 

6 

33 

2 

594 

35 

404 

22 

67 

4 

123 

9 

Vizze 

406 

23 

219 

13 

56 

2 

131 

8 

Aurina 

1,174 

47 

1,050 

21 

61 

3 

63 

3 

CroceAlta 

795 

37 

105 

5 

419 

17 

271 

15 

272 

23 

187 

21 

85 

2 

conceal  considerable  site  variations:  in  different  valleys 
of  equal  aspect  in  the  same  geographical  region,  plant 
density  may  vary  by  a  factor  of  three. 

As  far  as  individual  species  are  concerned,  the  highest 
concentration  of  cembran  pine  is  found  in  the  western  re- 
gion, with  736  trees  per  hectare  (excluding  Genova  where 
pine  is  absent).  This  average  is,  however,  influenced  by 
an  unusually  high  number  of  pines  in  Furva  (1,102  indi- 
viduals per  hectare),  without  which  cembran  pine  density 
would  have  been  closer  to  the  515  individuals  per  hectare 
recorded  in  the  southeast. 

Larch  dominates  the  shaded  slopes  in  the  west  (327  in- 
dividuals per  hectare),  but  is  less  important  in  the  south- 
east (114  individulals  per  hectare). 

The  sxmny  slopes  of  the  northeastern  sites  present  such 
a  great  variation  in  tree  density  that  averages  are  mean- 
ingless; and  shaded  slopes  were  monitored  in  one  area  only. 

Average  basal  area  (table  2)  revolves  aroimd  30  m^/ha, 
being  only  a  little  higher  on  sunny  slopes.  Values  show 
a  narrower  range  than  those  of  tree  nimibers  per  hectare, 
indicating  an  efficient  use  of  available  land. 

Diameter  Distribution 

As  shown  in  figure  2,  the  number  of  individuals  gener- 
ally decreases,  even  if  irregularly,  with  increasing  diam- 
eter class;  and  the  distributions  are  continuous  to  about 
65  cm  d.b.h.,  with  an  occasional  scatter  of  trees  in  size 
classes  85  to  110  cm. 

In  most  cases,  the  majority  of  cembran  pine,  larch, 
and  spruce  are  contained  in  the  size  classes  up  to  30  cm: 
(for  example,  Solda,  warm  slope:  80,  50,  and  75  percent; 
Badia,  cold  slope:  65,  60,  and  82  percent)  and,  quite  fre- 
quently, concentrated  solely  in  the  smallest  d.b.h.  class, 
to  10  cm  (for  example,  Viola,  warm  slope:  81,  59,  and  49 
percent;  Furva,  warm  slope:  53,  64,  and  75  percent). 

At  some  sites,  the  shape  of  the  distribution  shows  a 
"hump"  at  the  30  to  50  cm  d.b.h.  class  level:  as  on  the 
cold  slopes  of  Viola,  Genova,  and  Badia;  in  Fiemme;  and 
on  the  warm  slope  of  Croce  Alta.  This  affects  larch  and 
spruce  to  a  greater  extent  than  cembran  pine. 

Bigger  size  classes  contain  more  spruce  and  larch  than 
cembran  pine,  with  the  exception  of  some  large  pines  on 
north-facing  slopes. 


Looking  at  the  d.b.h.  distribution  in  relation  to  altitude, 
in  Badia  the  percentage  of  small-diameter  larches  rises 
with  increasing  altitude;  likewise,  the  number  of  big 
cembran  pine  trees  increases  and  that  of  large  spruces 
and  larches  diminishes.  In  Aurina,  the  number  of  small 
cembran  pines  increases  with  altitude,  with  spruce  and 
larch  disappearing  altogether  in  the  highest  belt. 

Age  Structure 

Since  age  structure  was  analyzed  only  in  some  of  the 
study  areas,  and  results  were  not  given  for  each  species 
separately,  figure  3  shows  the  "cored  ages"  of  all  trees,  in 
25-year  classes,  to  which  must  be  added  the  time  taken  to 
reach  1.3  m  height. 

Cembran  pine,  spruce,  and  larch  in  Viola  take  on  aver- 
age 15,  30,  and  40  years;  cembran  pine  in  Furva  20  and 
in  Funes  30  years  to  grow  to  1.3  m,  which  explains  the 
apparent  lack  of  trees  up  to  25  years  old  and  the  reduced 
percentage  of  26-  to  50-year-olds. 

But  even  so,  the  histograms  show  yoxmg  populations:  in 
Viola  (and  Furva),  79  percent  (86  percent)  of  the  trees  are 
yoxmger  than  100  years  and  only  7  percent  (6  percent)  are 
older  than  150  years.  In  Viola,  the  oldest  trees  found 
were  281  (larch),  274  (spruce),  and  150  (pine)  years  old, 
whereas  in  Furva  the  oldest  cored  tree  is  a  cembran  pine 
of  310  years;  the  oldest  larch  and  spruce  are  244  and  230 
years  old. 

In  Funes,  49  percent  of  trees  are  under  100  years  old, 
but  all  age  classes  to  300  years  are  well  represented.  The 
oldest  1  percent  of  trees  are  in  the  326  to  450  age  classes, 
including  a  larch,  a  cembran  pine,  and  a  spruce  at  431, 
396,  and  289  years  old. 

In  Aurina,  tree  populations  become  progressively 
younger  with  altitude.  Rising  from  the  lowest  to  the  high- 
est belt,  the  proportion  of  trees  under  100  years  old  is  57, 
74,  62,  and  91  percent.  Belt  3,  at  nearly  2,200  m  a.s.l., 
contains  the  oldest  trees,  with  17  and  5  percent,  respec- 
tively, in  the  201  to  250  and  251  to  300  age  classes. 

The  oldest  trees  stem-cored  were  at  Vizze:  a  557-year- 
old  larch  and  a  cembran  pine  of  443  years. 

A  proportional  relationship  between  age  and  diameter 
exists  only  within  very  wide  margins  so  that  tree  sizes  do 
not  allow  age  estimates;  thus,  a  25  cm  d.b.h.  Vizze  pine 


263 


200 


1-10    11-20    21-30    31-40    41-50   51-60  61-70 

DBH  (cm) 


200 

160- 


Val  Viola 

Shaded  slope 


1-10    11-20    21-30    31-40    41-50   51-60  61-70 

DBH  (cm) 


200 


160- 


Val  di  Solda 

Sunny  slope 


200 


1-10   11-20  21-30  31-40  41-50  51-60  61-70  71-80 

DBH  (cm) 


160- 


Val  di  Solda 

Shaded  slope 


1-10  11-20  21-30  31-40  41-50  51-60  61-70  71-80  81-90 

DBH  (cm) 


200 


Valfurva 

Sunny  slope 


1  1 — 

1-10    11-20    21-30    31-40    41-50   51-60  61-70 


200 


DBH  (cm) 


160- 


Valfurva 

Shaded  slope 


i    n     I  1  1  1  1 — 

1-10       21-30      41-50      61-70      81-90  101-110 


DBH  (cm) 


200 


160 


S.  8° 


40 


Val  Geneva 

Sunny  slope 


200- 


-T  1  r 

1-10  11-20  21-30  31-40  41-50  51-60  61-70  71-80  81-90 


160- 


2  80H 


40- 


Val  Genova 

Shaded  slope 


T — I — I — I — I — I — I — 

1-10       21-30      41-50      61-70      81-90  101-110 


DBH  (cm) 


DBH  (cm) 


200 


160- 


Val  di  Fiemme 

Sunny  slope 


1-10   11-20  21-30  31-40  41-50  51-60  61-70  71-80 

DBH  (cm) 


200 


160- 


Val  di  Fiemme 

Shaded  slope 


0     I     I    ^^^1—1     I     I      I     I  I 

1-10     21-30     41-50     61-70     81-90  111-120 

DBH  (cm) 


Figure  2 — Diameter  distribution  of  pine,  spruce,  and  larch. 


264 


200  ■ 
160- 


Val  Badia 

Shaded  slope 


1-10  11-20  21-30  31-40  41-50  51-60  61-70  71-80 

DBH  (cm) 


200- 


160- 


Monte  Croce  Alta 

Sunny  slope 


1-10    11-20    21-30    31-40    41-50   51-60  61-70 

DBH  (cm) 


200- 


160- 


Vai  Aurina 

1870-1 930m 


1-10      11-20      21-30      31-40     41-50  51-60 

DBH  (cm) 


200- 


160- 


Val  Aurina 

2190-2200m 


1-10      11-20      21-30      31-40     41-50  51-60 

DBH  (cm) 


200 


ISO  - 


Monte  Croce  Alta 

Shaded  slope 


1-10       21-30      41-50      61-70       81-90  101-110 

DBH  (cm) 


200 


160 


120 

e 
o 

ffl  80 


40 


Val  Aurina 

1980-2000m 


1-10  11-20  21-30  31-40 

DBH  (cm) 


200 


11-20  21-30 

DBH  (cm) 


Pinus  cembra 


Figure  2  (Con.) 


Picea  abies 


Larix  decidua 


265 


a> 


0)  20- 


0) 


0)  20- 


ABCDEFGHIJKLMN 


Age  (yrs) 


ABCDEFGH   I    J  KLMN 


Age  (yrs) 


50 


40- 


■g  30- 

a> 
u 

0)  20- 


10 


Val  di  Funes 


■g  30- 

u 

0)  20- 


ABCDEFGHIJKLMN 

Age  (yrs) 


A  =  0-25 
B  =  26-50 
C  =  51-75 
D  =  76-100 


E  =  101-125 
F  =  126-150 
G  =  151-175 
H  =  176-200 


AO      CD     EF     GH      IJ  KL 


Age  (yrs) 


I  =201-225 
J  =  226-250 
K  =  251 -275 
L  =  276-300 


M  =  301-325 
N  =  326-450 


Figure  3 — Age  distribution  of  pine,  spruce,  and  larch. 


can  be  from  60  to  250  years  old,  and  at  100  years  of  age 
its  d.b.h.  may  range  from  10  to  50  cm. 

From  the  d.b.h.  distribution,  the  "cored  age"  range  per 
species  for  each  10-cm  d.b.h.  class  and  the  approximate 
age  of  trees  at  1.3  m  "cored"  height,  stand  development  on 
the  shaded  Viola  slope  was  cautiously  reconstructed.  The 
oldest  larch  and  spruce  found  date  from  the  1660's  and 
1680's,  the  oldest  cembran  pines  only  from  the  1820's. 
The  1750's-1820's  saw  ample  larch  recruitment  parallel 
with  a  spruce  maximum  from  the  1780's-1850's,  still  testi- 
fied to  by  today's  peak  in  the  30-  and  40-cm  d.b.h.  classes. 
Larch  recruitment  culminated  in  the  1830's-1930's  when 
40  percent  of  today's  population  (to  10  cm  d.b.h.)  was  es- 
tablished; now  regeneration  is  continuing  at  a  slower 
pace.  Cembran  pine  recruitment  accelerated  after  the 
1850's.  Indeed,  51  percent  of  today's  population  (to  10  cm 
d.b.h.)  dates  from  1870-1970  and  regeneration  continues 
strongly,  while  spruce  regeneration  is  being  reduced. 

Regeneration 

Table  3  shows  that  plant  density  varies  considerably  be- 
tween study  areas  (from  111  to  1,646  individuals  per  hec- 
tare at  Geneva  and  Viola);  in  fact,  area  averages  conceal 


notable  variations  between  plots.  Also,  plant  density  is 
usually  higher  on  sunny  than  on  shaded  sites. 

Nearly  everywhere,  cembran  pine  regeneration  prevails 
over  that  of  spruce  and  larch  by  a  large  margin,  more  so 
on  warm  than  on  cold  slopes.  Spruce  definitely  prefers 
warm  sites. 

To  detect  another  indicator  of  stand  dynamics,  a  com- 
parison was  made  between  the  species  composition  of  the 
adult  stand  (table  1)  and  that  of  regeneration.  As  a  gen- 
eral trend,  the  dominance  of  cembran  pine  in  the  adult 
stand  is  being  consolidated  by  an  even  higher  proportion 
of  regeneration.  This  happens  in  Viola,  on  the  cold  slopes 
of  Solda,  Fiemme,  and  Croce  Alta,  in  Vizze  and  Aurina, 
and  on  the  simny  slope  of  Funes.  The  larch  wood  with 
cembran  pine  on  the  shaded  side  of  Furva  experiences 
increased  cembran  pine  regeneration,  and  the  sunny 
spruce-pine  stand  of  Fiemme  now  contains  more  than 
twice  as  many  small  cembran  pines  as  spruces. 

An  enquiry  dealing  with  regeneration  in  the  western 
region  (Bettini  1987)  showed  that  at  lower  altitudes 
(1,700  to  1,800  m)  cembran  pine  establishes  mainly  in 
large  clearings  within  the  wood.  It  does  best,  in  terms 
of  density,  at  higher  altitudes  (1,900  to  2,200  m)  where 
it  preferably  occupies  large  clearings,  open  spaces  just 
outside  the  woods,  and  open,  multistoried  woods. 


266 


Table  3 — Regeneration:  plant  density  (individuals  per  hectare)  and  species  composition  of  plants  >  =  0.2  m  <1 .3  m  height 


Location 

Sunny  slope 

Shaded  slope 

Total 

Pine 

Spruce 

Larch 

Total 

Pine 

Spruce 

Lard 

No. 

 Percent  -  -  - 

No. 

 Percent  -  - 

Viola 

1,646 

70 

9 

21 

745 

61 

7 

32 

Solda 

207 

74 

13 

13 

400 

68 

1 

31 

Furva 

673 

87 

8 

5 

518 

43 

0 

57 

Geneva 

111 

0 

38 

62 

128 

0 

20 

80 

Fiemme 

303 

67 

30 

3 

277 

89 

10 

1 

Badia 

312 

37 

26 

37 

Punas 

412 

87 

8 

5 

207 

64 

3 

33 

Vizze 

849 

80 

4 

16 

Aurina 

458 

88 

4 

8 

Croce 

1,249 

25 

24 

51 

459 

84 

6 

10 

Health  Conditions 

According  to  a  visual  examination  of  the  trees,  the  state 
of  health  of  the  timberline  woods  is  good.  On  average,  75 
percent  of  individuals  were  considered  healthy;  14  percent 
suffering;  3  percent  dying;  and  4/4  percent  standing/lying 
dead. 

The  majority  of  dead  and  dying  individuals  are  small 
trees,  vvdth  diameters  to  15  cm  (Genova)  or  30  cm 
(Fiemme),  v^hereas  the  proportion  of  mortality  involving 
big  trees  (above  55  cm  d.b.h.)  varies,  being  virtually  nil 
in  Fiemme  and  30  percent  of  deaths  in  Vizze. 

Generally,  deaths  of  yoimg  plants  are  due  to  competi- 
tion, tramphng,  browsing,  and  snow  fungi;  mortalities 
of  larger  trees  are  caused  by  lightning,  avalanches,  and 
landshdes. 

DISCUSSION 

The  stand  structure  of  mixed  cembran  pine,  European 
larch,  and  Norway  spruce  woodlands  at  the  upper  timber- 
line  is  extremely  diverse  on  a  local  scale.  Timberhne  lim- 
its, species  composition,  plant  density,  regeneration,  di- 
ameter distribution,  and  age  structure  vary  between 
adjacent  sites  of  the  same  valley  studied,  from  one  study 
area  to  another,  between  simny  and  shaded  slopes  of  the 
same  valley,  and  between  different  altitudes,  apparently 
as  a  result  both  of  physical  emironmental  factors  and  of 
himian  action,  frequently  hard  to  detect  and  explain. 

The  sequence  of  altitudinal  limits  (forest  limit  being 
succeeded  by  spruce  tree  line,  followed  by  larch  tree  line, 
topped  by  cembran  pine  tree  line)  is  somewhat  higher  on 
sunny  slopes  than  on  shaded  ones,  probably  due  to  more 
favorable  site  conditions  such  as  extended  insolation, 
higher  temperatures,  and  earlier  snow  melt. 

In  fact,  spruce  distribution  seems  to  be  strongly  limited 
by  heat  deficiency.  Larch  is  usually  more  numerous  on 
cold  sites  and  is  well  represented  at  all  altitudes  except 
for  the  highest  locations.  At  a  first  glance,  cembran  pine 
appears  to  have  no  preference  as  to  aspect.  However,  its 
performance  is  closely  linked  to  the  behavior  of  spruce 


and  larch,  as  cembran  pine  is  dominating  where  the 
latter  two  species  show  constraints.  Virtually  every- 
where, therefore,  cembran  pine  dominance  increases 
with  altitude. 

The  timberline  woods  are  predominantly  young  stands 
in  various  phases  of  development:  72  percent  of  the  trees 
are  younger  than  100  years;  23  percent  belong  to  the  101 
to  200  year  class;  4.75  percent  to  the  201  to  300  year 
class,  and  0.25  percent  are  over  300  years  old. 

This  age  structure  reflects  centuries  of  unabated  exploi- 
tation, which  culminated  in  the  last  centur>%  and  the  sub- 
sequent progressive  abandonment  of  the  mountains  with 
the  decline  of  agriculture,  which  has  given  rise  to  the 
spontaneous  recolonization  of  dereUct  pastures  and  de- 
graded woodlands. 

As  the  more  inhospitable,  shaded  slopes  were  aban- 
doned first,  woodland  returned  there  earlier,  as  shown  by 
a  larger  proportion  of  medium-sized  trees  on  cold  slopes. 

The  fact  that  bigger  d.b.h.  classes  contain  more  spruce 
and  larch  than  pine,  except  on  some  cold  slopes,  clearly 
reflects  the  traditional  practice  of  eradicating  cembran 
pine,  which  shaded  the  pasture.  Since  this  intervention 
ceased,  cembran  pine  has  been  able  to  reclaim  especially 
those  areas  with  reduced  competition  from  the  other  two 
species:  aspects  and  altitudes  too  cold  for  spruce,  and  wood- 
land stands  too  dense  for  larch  regeneration.  As  a  result, 
cembran  pine  now  dominates  the  young  generation  at  vir- 
tually all  sites  and  continues  to  colonize  high  altitudes. 

However,  tree  vegetation  is  favored  on  simny  aspects: 
benefiting  from  a  more  favorable  microclimate  and  greatly 
reduced  disturbance,  there  is  higher  tree  density  and 
more  regeneration  than  on  shaded  slopes. 

And  changes  are  still  under  way  at  timberline:  density 
is  increasing,  cembran  pine  percentage  is  rising,  woodland 
area  is  expanding  to  higher  elevations. 

It  is  assimied  that  the  present  forest  limit  is  lower  than 
the  potential  one.  It  is  not  clear  whether  all  the  area  be- 
tween forest  limit  and  tree  line  can  be  occupied  by  the 
new  stands,  whose  edges  are  sometimes  abrupt,  but  the 
irregular  ecotone  formed  by  scattered  trees  seems  to  be 
the  result  either  of  difficult  edaphic  conditions  or  of  hu- 
man activity  and  cannot  be  considered  an  expression  of 
climatic  conditions. 


267 


REFERENCES 

Bettini,  D.  1987.  La  rinnovazione  naturale  in  un  bosco  di 
Picea,  cembro  e  larice  nella  Valle  del  Torrente  Frodolfo 
(prov.  di  Sondrio).  University  of  Firenze.  109  p.  Thesis. 

Catalano,  L.  1987.  Studio  del  limite  del  bosco  nei 
complessi  di  Stores  e  Settsass  nell'alto  bacino  del  fiume 
Gadera  (prov.  BZ  e  BL).  University  of  Firenze.  152  p. 
Thesis. 

Favero  del,  R.;  Mas  de,  G.;  Lasen,  C.;  Paiero,  P.  1985.  II 

pino  cembro  nel  Veneto.  Regione  del  Veneto.  85  p. 
Filipello,  S.;  Sartori,  F.;  Vittadini,  M.  1980.  Le  associazioni 

del  Cembro  nel  versante  meridionale  dell'arco  alpino,  2. 

La  vegetazione:  aspetti  forestali.  Atti  dell'Istituto 

Botanico  e  Laboratorio  Crittogamico  dell'Universita'  di 

Padova.  14:  1-48. 
Hellweger,  V.  1989.  II  limite  superiore  del  bosco  e  degli 

alberi  e  le  condizioni  del  bosco  di  alta  montagna  a  Riva 

di  Tures  in  Val  Aurina  (BZ).  University  of  Firenze. 

170  p.  Thesis. 
Maniero,  R.  1987.  II  limite  superiore  del  bosco  e  degli 

alberi  in  Val  di  Fxmes  (BZ).  University  of  Firenze. 

127  p.  Thesis. 

Mayer,  H.;  Ott,  E.  1991.  Gebirgswaldbau  Schutzwaldpflege. 
Stuttgart:  Gustav  Fischer  Verlag.  587  p. 

Olivari,  M.  1983.  II  limite  superiore  del  bosco  e  degli 
alberi  in  alcune  zone  della  Val  di  Fiemme  (TN).  Univer- 
sity of  Firenze.  70  p.  Thesis. 


Pirelli,  P.  1984.  Composizione  e  struttura  del  bosco  in 

alta  montagna  in  Val  dei  Fomi,  Valfurva  e  Val  Zebru'. 

University  of  Firenze.  72  p.  Thesis. 
Piussi,  P.;  Schneider,  A.  1985. 1  limiti  superiori  del  bosco 

e  degli  alberi  in  Val  di  Vizze.  Annali  dell'Accademia 

Italiana  di  Scienze  Forestali.  34:  121-150. 
Piussi,  P.;  Proietti,  A.  M.  1986.  Struttura  e  rinnovazione 

del  pino  cembro  a  Monte  Croce  Alta.  Unpublished 

manuscript.  22  p. 
Rapella,  A.  1985.  II  limite  superiore  del  bosco  e  degli 

alberi  in  Val  Viola  (SO).  University  of  Firenze.  116  p. 

Thesis. 

Schiechtl,  H.  M.;  Stern,  R.  1975.  Die  Zirbe  (Pinus  cembra 
L.)  in  den  Ostalpen,  I.  Teil.  Angewandte 
Pflanzensoziologie.  23:  1-84. 

Schroeter,  C.  1908.  Das  Pflanzenleben  der  Alpen.  Zuerich: 
Albert  Raustein  Verlag.  806  p. 

Travan,  L.  1984.  Struttura  e  rinnovazione  nei  boschi  in 
alta  montagna  in  Val  di  Solda.  Unpublished  manu- 
script. 37  p. 

Unterrichter,  M.  1986.  II  Cirmolo.  Note  ecologiche, 
vivaistiche  e  selvicolturali.  Regione  Lombardia.  84  p. 

Zoanetti,  R.  1983.  II  limite  superiore  del  bosco  e  degli 
alberi  in  Val  Genova  (Trento).  University  of  Firenze. 
73  p.  Thesis. 


263 


CURRENT  DISTRIBUTION  OF  CEMBRA 
PINE  IN  THE  LECHTAL  ALPS 


Siegfried  Sauermoser 


Abstract — The  Lechtal  Alps  are  located  at  the  northern  rim 
of  the  Alps.  In  these  mountains  the  subalpine  forests  are  domi- 
nated by  spruce  (Picea  abies)  and  by  mountain  pine  (Pinus  mugo), 
while  only  a  few  cembra  pine  (Pinus  cembra)  stands  are  found 
there.  Cembra  pine  occurs  in  15  different  places,  each  of  them 
more  than  5  hectares. 


In  Austria,  we  have  the  so-called  forest-technical  service 
system  for  torrent  and  avalanche  control,  which  means 
that  we  are  not  only  responsible  for  regular  hydraulic 
engineering  and  technical  measures,  but  also  for  refores- 
tation of  torrent  catchments  or  avalanche  areas,  for  ex- 
ample, and  for  the  maintance  and  establishment  of  the 
protective  forests  in  our  Country.  The  forest-technical 
system  of  torrent  and  avalanche  control  originated  in 
France  and  was  introduced  to  Austria  during  the  middle 
of  the  last  century. 

The  forest-technical  service  has  replanted  about  2,500 
hectares  of  timberline  area  during  the  last  30  years,  mainly 
using  spruce  (Picea  abies),  larch  (Larix  decidua),  and 
cembra  pine  (Pinus  cembra). 

Cembra  pine  is  an  especially  important  tree  for  affores- 
tation in  protective  forests.  Being  an  evergreen  conifer,  it 
possesses  the  highest  snow  interception  (Aulitzky  1982), 
which  is  extremely  important  for  reforestation  in  ava- 
lanche areas.  In  addition,  cembra  pine  is  very  long-lived 
and  is  able  to  form  stable  forest  stands  (Mayer  1977). 
Moreover,  it  is  the  most  frost-resistant  subalpine  tree 
(Tranquillini  1963). 

Cembra  pine  grows  well  on  crystalline  substratimi  in 
the  central  alpine  area,  where  it  is  primarily  used.  The 
situation  is  quite  different  at  the  northern  rim  of  the  lime- 
stone mountains.  Cembra  pine  is  not  as  common  there, 
and  reforestation  with  cembra  pine  must  be  more  care- 
fully related  to  site  conditions  than  in  the  central  Alps. 
In  the  limestone  Alps  cembra  pine  occurs  only  in  a  very 
small  area  and  is  not  important  to  the  forestry  industry. 
Thus,  there  are  only  a  few  studies  on  it,  although  cembra 
pine  is  very  important  for  the  protective  forests.  The  only 
study  about  silvicultural  treatment  and  natiu-al  regenera- 
tion in  a  cembra  pine-larch  forest  of  the  northern  lime- 
stone moimtains  was  done  by  Kleine  (1984). 

The  following  questions  would  be  of  special  interest  to 
the  expert: 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Seigfried  Sauermoser  is  Engineer  of  the  Technical  Forest  Service  for 
Torrent  and  Avalanche  Control,  6600  Lechaschau,  Austria. 


•  How  do  the  present  cembra  pine  stands  develop,  how 
old  are  they,  what  are  their  degree  of  stability  and 
the  potential  for  regeneration? 

•  How  do  these  cembra  pine  stands  develop  under  the 
influence  of  the  decline  of  grazing,  on  the  one  hand, 
and  the  high  game  population  on  the  other? 

•  How  much  was  the  distribution  of  cembra  pine  influ- 
enced by  himian  activities,  and  what  was  the  original 
distribution? 

•  And  the  most  important  question:  on  which  sites  can 
cembra  pine  be  successfully  used  for  reforestation? 

However,  due  to  the  fact  that  our  research  did  not  begin 
imtil  last  year,  I  cannot  answer  these  questions  yet;  but  I 
will  try  to  give  you  a  siirvey  of  the  present  cembra  pine 
stands  in  the  Lechtal  Alps. 

LOCATION 

The  Lechtal  Alps  are  located  at  the  northern  rim  of  the 
limestone  Alps  between  the  10th  and  11th  longitudinal 
degree  (east  of  Greenwich)  and  between  the  47th  and  48th 
latitudinal  degree  (fig.  1).  From  east  to  west  they  extend 
for  about  60  kilometers  and  for  about  20  kilometers  in  the 
north-south  direction.  The  total  area  is  about  1,000  km'^. 

CLIMATE 

The  location  on  the  northern  border  of  the  Alps  results 
in  a  high  precipitation  level  in  summer  as  well  as  in  win- 
ter. The  average  precipitation  level  along  the  Lechtal  is 


Figure  1 — Lechtal  Alps  location. 


269 


between  1,200  mm  and  1,400  mm  per  year  at  an  elevation 
of  1,000  m.  Air  currents  from  the  west  and  northwest 
bring  moist  air  from  the  Atlantic  regions,  and,  when  forc- 
ed crossing  the  mountains,  frequent  rain  falls  in  this  area. 

From  the  climatological  point  of  view,  the  position  of 
plant  communities  can  be  described  by  the  angle  resulting 
from  the  local  amount  of  precipitation  drawn  versus  el- 
evation of  the  area.  Areas  that  woiild  be  located  in  the 
sector  >45  degrees  are  considered  to  have  a  continental 
climate  (=  "hygrische  Kontinentalitat,"  Gams  1931).  The 
"hygrische  Kontinentalitat"  of  the  Lechtal  Alps  lies  be- 
tween 40  and  50  degrees  (fig.  2).  The  central-alpine  re- 
gions, main  area  of  larch-cembra  pine  forests,  are  more 
continental  (70-90  degrees). 

GEOLOGY 

Geologically,  the  Lechtal  Alps  are  very  complicated. 
The  landscape  is  primarily  determined  by  two  moimtain 
ranges.  The  main  mountain  range  is  dolomite,  which 
forms  rugged,  sterile  mountains.  In  these  dolomite  areas, 
the  most  common  soil  is  redsina,  with  differing  depths  of 
humus  layer.  The  southern  slopes  are  particularly  gentle 
and  dry,  due  to  easy  infiltration  of  precipitation  water. 


On  the  other  hand,  there  are  mountains  that  are  com- 
posed of  marl,  that  is  more  susceptible  to  weathering  and 
erosion.  These  areas,  characterized  by  a  more  smooth  to- 
pography, are  covered  with  vegetation  up  to  the  mountain 
tops.  The  soil  is  deeper,  compared  to  the  dolomite  moun- 
tains, and  rich  in  clay,  while  the  humus  layer  is  shal- 
lower. In  these  mountains,  the  effects  of  anthropogenic 
influence  are  especially  evident  as  they  have  been  cleared 
and  cultivated  for  centuries  (fig.  3). 

VEGETATION  AND  PLANT 
COMMUNITIES 

The  valleys  of  the  Lechtal  Alps  are  located  at  an  eleva- 
tion of  about  1,000  m,  and  the  highest  peaks  are  3,000  m 
high.  Due  to  elevation  and  climatic  conditions,  which  are 
strongly  influenced  by  air  currents  from  the  Atlantic,  the 
Lechtal  Alps  are  characterized  by  typical  north  alpine 
vegetation.  Spruce  and  spruce-fir  (Abies  alba)  forests 
dominate  up  to  an  elevation  of  about  1,500  m,  especially 
on  the  northern  slopes. 

The  timberline  is  located  at  about  1,900  m,  and  is 
formed  by  different  tree  species.  Mountain  pine  (Pinus 
mugo)  prevails  and  forms  large  stands  in  the  dry  dolomite 


4C/0 


10* 


500 


WOO 


Larch  -  cembra  forest 
Larch  -  spruce  forest 


J  L 


1500 


20C0  mm 


treeline 


Figure  2— "Hygrische  Kontinentalitat"  according  to 
Gams  (1931)  for  different  forest  types  and  geographi- 
cal regions.  The  points  1  -1 3  show  different  sites  of  the 
Lechtal  Alps.  The  points  1 ,  2,  and  3  are  on  the  south- 
ern rim  (more  continental);  the  points  10  and  1 1  on  the 
north  rim  indicate  climatic  conditions  more  influenced 
by  air  currents  from  the  Atlantic. 


270 


Figure  3 — Typical  mountains  composed  of  marl 
in  the  Lechtal  Alps.  The  clearcut  slopes  have 
been  used  for  agriculture  since  the  14th  century. 
After  the  strong  decline  of  livestock  farming  the 
slopes  will  be  reforested  to  create  protective 
forests. 


areas.  Spruce  and  fir  or  larch  appear  only  singly  here. 
This  type  of  forest,  including  large  stands  of  dwarf  moxm- 
tain  pine  with  single  spruce  or  larch  trees,  is  found  prima- 
rily on  southern  slopes.  In  marl  areas,  the  forest  limit  is 
mostly  formed  by  spruce.  On  moist  northern  slopes  it  is 
joined  by  green  alder  (Alnus  viridis). 

Where  can  cembra  pine  be  found  in  the  Lechtal  Alps? 

Cembra  pine  occurs  in  about  20  different  places,  each 
with  a  coverage  of  more  than  5  ha.  The  largest  stand  of 
cembra  pine  is  about  20  ha.  In  many  other  places,  cembra 
pine  occurs  singly  in  contact  with  other  tree  species. 

When  comparing  the  distribution  pattern  of  cembra 
pine  with  the  geological  map,  it  becomes  obvious  that 
cembra  pine  grows  primarily  on  dolomite  and  limestone, 
but  is  less  frequent  on  marl.  There  also  is  a  relationship 
of  cembra  pine  range  to  soil  conditions  as  it  occurs  in  con- 
tact areas  between  dolomite  and  marl.  Haupt  (1983),  in 
his  study  on  the  forest  commiuiities  of  the  Lechtal  Alps, 


also  observed  this  phenomenon.  Similar  observations 
were  made  farther  east  in  the  Karwendel  Mountains 
by  Vareschi  (1931).  Vareschi  found  the  occurrences  of 
cembra  pine  in  the  upper  Isar  Valley  to  be  related  to  lay- 
ers of  dolomite  and  raibler.  Raibler  stratum  is  a  geologi- 
cal formation  consisting  of  different  limestones  and  marls. 

Although  mainly  marl  areas  were  cleared  and  settled, 
this  cannot  be  the  only  reason  for  the  absence  of  stone 
pine  on  marl.  I  believe  germination  conditions  to  be  worse 
on  marl  than  on  dolomite,  because  of  the  shallow  humus 
layer,  smooth  topography  with  long-lasting  snow  cover, 
and  stronger  vegetation  competition  on  the  nutrient-rich 
soils.  If  stone  pine  is  artificially  introduced  to  marl  areas, 
it  shows  great  fertility,  which  might  support  the  theory 
that  cembra  pine  is  less  fi'equent  on  marl  because  of  its 
low  competition  ability.  Moreover,  as  cembra  pine  seeds 
germinate  most  successfully  at  moisture  conditions  of 
60-70  percent  and  temperatures  between  20  and  22  °C 


271 


Figure  A — Southwest-exposed  plateau,  the  so- 
called  "cembra  pine-flatland"  with  cembra  plne- 
spruce-mountain  pine  forest. 


(Nather  1958),  marl  soils  might  be  considered  to  be  too 
cold  to  allow  successful  regeneration. 

The  situation  is  different  on  redsina,  which  can  occur 
on  dolomite.  Cembra  pine  can  be  found  on  northeast  and 
northwest  knolls  and  ridges,  where,  due  to  low  erosion, 
high  raw  humus  layers  can  develop.  Therefore,  cembra 
pine  is  primarily  found  on  knolls  and  crests  that  are  north- 
east or  northwest  exposed.  There  is  only  one  larger  stone 
pine  stand  on  south-  or  southwest-exposed  slopes.  It  is 
called  the  "cembra  pine  flat  land"  and,  as  the  name  indi- 
cates, it  is  a  plateau  at  an  altitude  of  about  1,800  m  (fig.  4). 
The  other  cembra  pine  stands  are  located  between  1,300 
and  1,950  m. 

At  lower  elevations,  cembra  pine  often  occurs  associated 
with  fir,  mountain  maple  (Acer  pseudoplatanus),  or  beech. 
These  are  contact  areas  for  mixed  forests,  which  are  also 
foimd  primarily  on  northern  exposures.  Cembra  pine  and 
spruce  are  often  found  together,  and  contact  to  the  subal- 
pine  spruce  forests  is  very  common. 

In  the  Lechtal  Alps,  larch  and  cembra  pine  are  seldom 
foimd  together.  There  is  only  one  forest  where  the  number 
of  larch  trees  exceeds  cembra  pines,  and  where  regenera- 
tion of  larch  is  abundant.  Contact  with  the  spruce  is 
much  more  common  in  the  Lechtal  Alps  (fig.  5).  Without 
pollen  analysis  studies,  it  is  difficult  to  guess  how  much 
forest  communities  are  anthropogenically  influenced,  or 
whether  this  phenomenon  is  simply  the  result  of  natural 
forest  development  on  carbonate  ground.  Mayer  (1974) 
described  three  varieties  of  "carbonate-larch-cembra  pine 
forests"  (Larici-Cembretum-Rhododendretosum  hirsuti). 
He  differentiated  between  the  tjT)ical  form  on  plateau  ar- 
eas, the  high  shrub-green  alder  variety  occurring  on  lias- 
dogger  substratum,  and  the  mountain  pine  variety.  The 
Lechtal  Alps'  stone  pine  forests  are  for  the  most  part  to 
be  attributed  to  the  dwarf  mountain  pine  variety.  The 
transition  to  pure  dwarf  mountain  pine  forests  is  fluid 
and  cannot  be  sharply  defined. 


STRUCTURE  AND  AGE 

Cembra  pine  generally  stands  very  loosely  and  is  con- 
centrated only  in  a  very  small  area  on  exposed  moimtain 
ridges.  At  24  sites,  the  maximimi  age  of  cembra  pine  was 
determined  by  tree-ring  dating.  At  each  site,  the  visually 
oldest  tree  was  examined.  The  tree  ages  ranged  from  150 
to  600  years.  The  apparently  oldest  cembra  pine  dated 
has  a  diameter  of  111  cm  and  is  probably  about  600  years 
old.  An  exact  age  determination  was  not  possible  because 
the  tnmk  is  rotten.  Thus,  the  age  was  only  estimated,  re- 
ferring to  comparable  trees.  In  the  group  containing  this 
old  cembra  pine,  there  are  four  other  cembra  pine  trees, 
in  the  middle  of  a  stand  of  dwarf  mountain  pine.  These 
stone  pines  are  about  330,  290,  190,  and  150  years  old. 
These  age  differences  in  cembra  pine  stands,  which  became 
apparent  from  the  increment  cores,  cannot  be  recognized 
with  the  naked  eye,  as  the  trunks  look  more  and  more  the 
same  when  getting  older. 

REGENERATION 

Regeneration  is  primarily  found,  as  in  the  central  al- 
pine regions  (Oswald  1963),  on  sites  with  early  snowmelt 
such  as  small  hummocks.  They  are  usually  built  up  by 
organic  matter  originating  fi"om  moss  and  dwarf-shrub 
vegetation.  The  organic  layer  is  approximately  30  to  40  cm 
thick.  It  is  separated  from  the  dolomite  bedrock  by  a  nsir- 
row  weathered  horizon.  The  sites  are  covered  with  Pinus 
mugo.  Rhododendron  ferrugineum,  Vaccinium  myrtillus, 
Vaccinium  vitis-idaea,  Calamagrostis  villosa,  and  Homo- 
gyne  alpina.  The  dominating  mosses  are  Hylocomium 
splendens,  Dicranum  scoparium,  Polytrichum  formosum, 
Rhytidiadelphus  triquetrus,  and  Pleurozium  schreberi. 
On  the  dwarf-shrub  hummocks  there  are  acidophilic 
plants,  while  between  and  around  the  small  hills  there 
are  more  calcicolous  plants  like  Calamagrostis  uaria, 
Sesleria  varia,  Carex  sempervirens,  Carex  firma,  Rhodo- 
dendron hirsutum,  and  Carex  ferrugineum  (fig.  6). 


272 


273 


Apparently,  the  decline  of  alpine  grazing  had  a  positive 
effect  on  regeneration  of  cembra  pine.  Although  exact 
studies  are  still  needed,  cembra  pine  regeneration  occurs 
especially  in  the  areas  of  former  alpine  pastures.  Most  of 
the  plants  are  up  to  30  cm  high.  Larger  plants  are  rare. 
As  previously  mentioned,  this  could  be  explained  by  re- 
generation of  cembra  pine  during  the  last  30  years  (since 
the  decline  of  grazing).  However,  it  may  also  be  possible 
that  cembra  pine  is  not  able  to  grow  higher  imder  the 
present  climate. 

In  this  connection,  the  damage  done  by  game  to  cembra 
pine  should  be  mentioned.  In  our  afforestations  game 
does  harm  mainly  to  cembra  pine.  Unfortimately,  in  the 
Lechtal  Alps  decline  of  alpine  grazing  runs  parallel  with 
an  increase  in  game.  So  today,  game  density  is  very  high 
in  this  area.  Only  exact  studies  will  be  able  to  determine 
the  degree  of  damage  caused  to  the  artificial  and  natural 
regeneration  of  cembra  pine. 

I  hope  to  have  given  you  a  small  survey  on  the  cembra 
pine  occurrences  in  the  Lechtal  Alps.  However,  additional 
information  can  only  be  expected  from  more  exact  studies 
in  the  future. 

REFERENCES 

Aulitzky,  Herbert;  Turner,  Hans.  1982.  Bioklimatische 
Grundlagen  einer  standortsgemaPen  Bewirtschaftung 


des  subalpinen  Larchen-Arvenwaldes.  Mitt.  Eidg.  Anst. 

f.  Forstl.  Versuchswesen.  58(2).  579  p. 
Gams,  Helmut.  1931.  Die  klimatische  Begrenzung  von 

Pflanzenarealen  und  die  Verteilung  der  hygrischen 

Kontinentalitat  in  den  Alpen.  Zeitschr.  der  Ges.  fur 

Erdkunde  zu  BerUn.  9/10:  331-346. 
Haupt,  Wolfgang.  1983.  Die  aktuelle  Vegetation  der 

ostlichen  Lechtaler  Alpen.  I.  Waldgesellschaften.  Veroff. 

Musetmi  Ferdinandevun  Innsbruck  Bd.  3.  67  p. 
Kleine,  Michael.  1984.  Waldbauliche  Untersuchungen  im 

Karbonat-Larchen-Zirbenwald  Warscheneck/Totes 

Gebirge  mit  Verkarstungsgefahr.  Diss.  Boku.  150  p. 
Mayer,  Hannes.  1974.  Walder  des  Ostalpenraumes. 

Gustav  Fischer  Verlag,  Stuttgart.  344  p. 
Mayer,  Hannes.  1977.  Waldbau.  Gustav  Fischer  Verlag, 

Stuttgart.  483  p. 
Nather,  Hans.  1958.  Zur  Keimung  der  Zirbensamen. 

Centralblatt  ges.  Forstwesen.  75:  61-70. 
Oswald,  Helfried.  1963.  Verteilimg  und  Zuwachs  der 

Zirbe  (Pinus  cembra  L.)  der  subalpinen  Stufe  an  einem 

zentralalpinen  Standort.  Mitt.  Forstl.  Bundesversuch- 

sanstalt  Mariabrunn.  60:  437-500. 
Tranquillini,  Walter.  1963.  Uber  die  Frostresistenz  der 

Zirbe.  Mitt.  Forstl.  Bimdesversuchanstalt  Mariabrunn. 

60:  547-562. 

Vareschi,  Volkmar.  1931.  Die  Geholztypen  des  obersten 
Isartales.  Universitatsverlag  Wagner,  Innsbruck. 


274 


GEOGRAPHICAL  DIFFERENTIATION 
AND  DYNAMICS  OF  SIBERIAN  STONE 
PINE  FORESTS  IN  EURASIA 

E.  P.  Smolonogov 


Abstract — The  study  reported  here  is  based  on  comprehensive 
long-term  investigations  (1960-90).  Regularities  in  geographical 
differentiation  and  biogeocenotic  pol)Tnorhism  of  stone  pine  for- 
ests as  well  as  rhythmic  changes  in  forest  composition  and  forest 
structure  dependent  on  reproduction  and  age  dynamics  are  con- 
sidered. Based  on  these  ecological  aspects  a  system  that  allows 
control  of  the  structural  development  of  the  coenosis  (morpho- 
coenogenesis)  and  a  strategy  for  use,  restocking,  and  preserva- 
tion of  stone  pine  forests  were  developed.  Stone  pine  forests  are 
the  most  essential  landscape  structures  of  the  Ural-Siberian 


In  Eurasia,  forests  that  include  or  are  dominated  by 
Siberian  stone  pine  extend  over  almost  70  longitudinal 
degrees  from  the  western  foothills  of  the  Urals  to  the  up- 
per drainage  area  of  the  Aldan  River  in  eastern  Siberia 
and  over  22  latitudinal  degrees  from  north  to  south,  from 
the  lower  Enisei  River  to  the  upper  drainage  area  of 
Orkhon  River.  The  total  area  covered  by  the  stone  pine 
forests  amounts  to  40  million  hectares.  Over  95  percent 
of  the  forest  area  belongs  to  Russia,  the  rest  to  northern 
Mongolia. 

The  formation  of  modern  ranges  of  stone  pines  and  pro- 
pagation of  forests  with  stone  pines  in  Eurasia  was  influ- 
enced by  mviltiple  intermittent  glaciations.  It  is  supposed 
that  stone  pines  had  a  continuous  range  before  the  glacial 
epoch,  while  during  cold  periods  they  survived  only  in  some 
upland  refugia  in  Europe,  in  the  Urals,  in  Siberia,  and  in 
central  south-eastern  Asia  (Kolesnikov  1956;  Luzganov 
and  Aboimov  1977;  Neishtadt  1957;  Nepomilueva  1974; 
Tolmachev  1954;  and  others).  This  might  explain  why 
Swiss  and  Siberian  stone  pines  became  separate  species 
(Bobrov  1978). 

A  TWO-PART  RANGE 

B.  P.  Kolesnikov  (1966)  considers  Siberian  stone  pine 
a  developing  species  that  still  is  extending.  In  fact,  comp- 
ared to  other  stone  pines,  it  has  the  widest  range.  Sibe- 
rian stone  pine  grows  and  propagates  well  under  severe 
highland  conditions  of  the  Urals,  southern  Siberia,  and 
central  Asia,  but  stone  pine  forests  also  occupy  large  areas 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 


E.  P.  Smolonogov  is  Professor  of  Forestry,  Institute  of  Forest,  Ural  Divi- 
sion of  Russian  Academy  of  Sciences,  Bilimbayevskaya  St.  32-A,  620134 
Ekaterinburg,  Russia. 


on  swampy  fluvioglacial  and  alluvial  plains  of  western  Si- 
beria. Stone  pine  that  immigrated  from  the  southern  re- 
gions to  western  Siberia  did  not  occur  there  before  10,000 
to  20,000  years  ago. 

Hence,  with  respect  to  this  history,  the  range  of  this 
species  should  be  divided  into  two  parts:  the  younger 
forested  part  of  the  Urals,  western  and  eastern  Siberia, 
where  migration  of  the  stone  pine  and  other  forest-forming 
species  followed  deglaciation,  and,  on  the  other  hand,  the 
ancient  part  of  the  mountain  systems  of  southern  Siberia 
and  central  Asia,  where  stone  pine  has  been  existing  for 
many  millions  of  years,  forming  stable  moimtain-forest 
flora  complexes. 

Also,  the  plains  of  western  and  eastern  Siberia  are 
not  homogeneous.  From  north  to  south  they  are  divided 
into  latitudinal  forest  zones  and  subzones  (Krylov  1961; 
Smolonogov  and  others  1970),  which  are  different  in  total 
duration  of  forest-,  swamp-,  and  soil-forming  processes 
(from  2,500  to  3,000  to  10,000  to  12,000  years),  thermal 
and  energy  conditions,  and  character  of  biosphere  pro- 
cesses. The  differences  are  great.  The  northern  forests 
are  mainly  fovmd  on  continuous  permafrost  soils,  thus 
growing  under  present  periglacial  conditions,  while  the 
southern  and  middle  forest  zones  are  located  within  the 
Eurasian  temperate  belt,  characterized  by  hydrothermal 
or  ecological  conditions  most  favorable  to  forest  growth. 

The  wide  range  of  Siberian  stone  pine  reflects  a  better 
ecological  and  geographical  adaptability  compared  to 
other  stone  pines  and  many  associated  forest-forming 
species.  Hence,  in  different  parts  of  its  range  stone  pine 
and  its  associated  species  form  forest  communities  that 
are  different  in  terms  of  species  composition  (depending 
on  the  landscape  structure),  specific  natural  and  forest- 
site  conditions,  chorology,  and  many  other  properties. 

REGIONAL  DIFFERENTIATION 

These  spatial  and  geographical  differences  can  be  de- 
scribed as  follows.  The  numbers  refer  to  figure  1. 

1.  Far  north:  Most  common  forest  communities  are  pre- 
timdra,  open  larch-spruce. 

2.  Ural-Siberian:  Larch-spruce-stone  pine  forests  in 
the  Ural  Mountains  and  western  Siberia  (Larix  sibirica 
Lebed.)  and  in  eastern  Siberia  Larix  dahurica  Turcz.  and 
Picea  obovata  Lebed.,  including  mountain  birch  {Betula 
tortuosa  Lebed.),  and  white  birch  {Betula  pubescens  L.). 

3.  Northern  taiga  subzone:  Spruce-stone  pine  forests 
with  an  admixture  of  fir  {Abies  sibirica  Lebed.),  white 
birch  {Betula  pubescens  L.),  drooping  birch  {Betula  ???) 


275 


dominating  early  phases  of  forest  reestablishment  and 
development,  are  very  common. 

4.  Middle  and  southern  taiga  subzones:  Fir-spruce- 
stone  pine  forests  with  the  same  birch  species  and  aspen 
(at  early  successional  phases)  as  the  northern  taiga. 

5.  Foothills  and  on  low  mountains  of  southern  Siberia: 
Fir-stone  pine  forests  with  an  admixture  of  spruce  com- 
monly occur.  Birch  species  and  aspen,  same  as  those  in 
the  taiga,  are  typical  during  early  phases  of  restoration 
and  forest  development. 

6.  Northeast:  The  stone  pine  range  is  dominated  by 
specific  light-demanding  conifers  of  stone  pine,  Dahurian 
larch,  and  Scots  pine  (Pinus  sylvestris  L.).  Asian  white 
birch  (Betula  platyphylla  Sukacz.)  occurs  at  early  phases 
of  forest  succession. 

7.  Highlands  of  southern  Siberia  and  the  northern 
macroslope  of  the  central  Asian  mountains  (northern 


Mongolia):  Larch-stone  pine  and  piure  stone  pine  forests 
occur  with  some  admixture  of  spruces  and  white  birches 
in  the  river  valleys  of  the  low  mountains. 

Natural  conditions  of  the  regional  varieties  of  stone  pine 
forests  vary  from  region  to  region,  but  the  variances  can  be 
explained  by  the  differences  in  vegetation  history  and  post- 
glacial landscape  development,  and  thus  in  the  character 
and  intensity  of  biogeocenotic,  soil-,  and  forest-formation 
processes.  Naturally,  the  pathways  of  forest  development 
and  temporal  changes  in  ph3^ocenoses  also  vary. 

REPRODUCTION  AND  AGE 
DYNAMICS 

The  most  important  forms  of  temporally  changing  forest 
commimities  can  be  described  as  follows: 


Figure  1 — The  range  of  Siberian  stone  pine  and  spread  of  forests  with  its  admixture.  1 — Far 
North:  pretundra  open  larch-spruce  forests;  2 — Ural-Siberian:  larch-spruce-stone  pine  forests 
(western  Siberia:  Larix  sibirica  Lebed.;  eastern  Siberia:  Larix  dahurica  Turcz.,  Picea  obovata 
Lebed.)  including  mountain  birch  {Betula  tortuosa  Lebed.)  and  white  birch  (Betula  pubescens  L.); 
3 — Northern  taiga  subzone:  spruce-stone  pine  forests  with  additional  fir  (Abies  sibirica  Lebed.), 
white  birch  (Betula  pubescens  L.)  and  drooping  birch  and,  sometimes,  aspen  (Populus  tremu- 
loides  L.)  dominating  at  early  phases  of  forest  stand  and  restoration,  are  very  common;  4 — Middle 
and  southern  taiga  subzone:  fir-spruce-stone  pine  forests  with  the  same  birch  species  and  aspen 
(at  early  successional  phases)  prevail;  5 — Foothills  and  low  mountains  of  southern  Siberia:  fir- 
stone  pine  forests  including  spruce,  birch  species,  and  aspen;  6 — Northeast:  light-demanding 
conifer-stone  pine  forests  with  Dahurian  larch  and  Scots  pine  (Pinus  sylvestris  L.).  Asian  white 
birch  (Betula  platyphylla  Sukacz.)  occurs  at  early  phases  of  forest  succession;  7 — Highlands  of 
southern  Siberia  and  central  Asian  mountains:  larch-stone  pine  and  pure  stone  pine  forests  with 
some  additional  spruces  and  white  birches  in  the  river  valleys  of  low  mountains  grow  there. 


276 


80 


® 

(D 

<i» 
E 

O 
> 


Period  1 


phase  1   ,       phase  2 


I 


1\ 


Period  2 


phase  3      phase  4 


■  -  -  Period  3   

phase  5  j         phase  6 

I 

Stone  Pine   

Spruce-Fir  

Birch-Aspen  

Spruce-Fir  

Second  Generation 


Age  (years) 


Figure  2 — Reproduction-age  dynamics  scheme  of  modal  stands  in  midtaiga, 
moss-herbaceous,  hillside  stone  pine  forests  of  the  Trans-Urals  submontane- 
plain  forest  site  province.  Stand  admixture  variation:  1 — Siberian  stone  pine; 
2 — dark-coniferous  species  (Siberian  spruce  and  fir);  3 — deciduous  species 
(birch,  aspen);  4 — second  generations  of  dark-coniferous  species. 


•  Age  dynamics  that  characterize  variation  in  phyto- 
cenoses  (including  alternation  of  age  classes  of  woody 
plants)  subject  to  a  weak  influence  of  external  dam- 
aging factors  in  the  course  of  long-term  existence  of 
the  forest  communities.  Probably,  climax  forest  com- 
mimities  can  be  formed. 

•  Reproduction  and  age-dynamics  after  fires  that  com- 
pletely destroyed  the  forest  on  large  areas.  In  this 
case  forest  succession  starts  again,  and  relatively 
even-aged  stands  are  formed. 

•  Reproduction  and  age-dynamics  after  clearcut.  For- 
est succession  starts  again,  but  now  includes  compo- 
nents of  previous  phytocenoses  and  some  elements 
resulting  from  silvicultural  management.  A  rela- 
tively even-aged  stand  is  formed. 

The  first  type  of  forest  dynamics  is  typical  of  moist  sites 
at  low  elevation,  in  depressions,  and  on  hillsides  moist- 
ened by  outcropping  water.  The  second  type  occurs  all 
over  the  regions  and  is  typical  of  fi^esh  and  dry  sites. 
Stone  pine  forests  of  the  Ural  Mountains,  Siberia,  and 
central  Asia  were  "fire-bom."  The  third  type  is  foimd  in 
regions  of  intensive  forest  exploitation. 

The  successional  behavior,  formation,  and  temporal 
variation  of  stone  pine  forests  in  different  regions  have 
been  intensively  studied  (Smolonogov  1990).  Tree  mensu- 
ration data  were  processed  by  the  probabilistic-statistical 
method.  The  method  derives  biometric  stand  characteris- 
tics related  to  age  classes,  which  are  presented  by  computed 
growth  curves  for  modal  communities  of  each  forest  type 
studied.  The  "Ural  version"  of  the  genetic  classification 


(Kolesnikov  and  others  1974;  Smolonogov  1990)  served 
as  the  topological  basis.  It  well  reflects  the  altitudinal- 
orographic  and  latitudinal-zonal  differentiation  of  forest 
sites  and  forest  cover,  while  the  forest  type  includes  both 
natiu-al  and  derivative  (potential  natviral)  forest  commu- 
nities growing  in  the  same  forest-site  types.  The  stand  is 
imeven  aged,  and  forms  the  same  natural-genetic  develop- 
ment series.  This  approach  to  studying  the  dynamics  is 
supported  by  the  well-known  concept  that  space  succes- 
sion may  correspond  to  temporal  succession  of  communi- 
ties (Clements  1928). 

The  investigations  show  that  a  complete  cycle  of  changes 
occurring  during  succession  of  forest  commvmities  that 
appear  to  have  been  heavily  damaged  is  usually  divided 
into  three  periods.  The  periods  are  subdivided  into  phases 
that  follow  each  other  in  time  and  reflect  quantitative  and 
qualitative  rhythms  of  the  forest  community  development 
(morphocenogenesis).  The  age  cycle  of  changes  in  each 
generation  of  commimity-forming  ligneous  plants  can  be 
broken  down  into  ontocenogenesis  stages. 

A  generalized  reproduction-age  dynamics  scheme  of  one 
of  the  types  of  mountain  stone  pine  forests  of  the  northern 
Urals  is  shown  in  figure  2: 

First  Period — Restoration  and  formation  of  stone  pine 
forests  dominated  by  rapidly  growing  deciduous  species 
(birch,  aspen).  Stone  pine  and  spruce  and  fir  (the  latter 
ones  are  species  of  the  dark  taiga)  are  growing  in  the 
lower  stories  of  the  biogeocenoses.  Potential  stone  pine- 
deciduous  commvmities  (actually  yovmg  stone  pine  growth) 
are  formed  in  this  period.  The  period  lasts  for  80  to  100 


277 


years.  It  is  subdivided  into  two  phases.  The  first  phase 
(1)  is  characterized  by  natural  reproduction  and  formation 
of  deciduous  young  growth  and  stone  pine  undergrowth. 
During  the  second  phase  (2)  the  upper  deciduous  story  is 
stabiHzed  and  its  degradation  begins.  Spruce,  fir,  and 
stonepine  grow  in  the  understories  of  the  communities. 

Second  Period — Formation  of  stone  pine  forests  sub- 
ject to  the  influence  of  prevailing  and  forest-forming 
spruce  and  fir.  Stone  pines  form  the  second  layer,  while 
rapidly  growing  trees  prevail  in  the  upper  story,  where 
stone  pines  may  account  for  up  to  20  to  30  percent  of  the 
growing  stock.  During  this  period  taiga  develops  that  is 
a  potential  stone  pine  forest  (actually  a  medium-age  stone 
pine  forest).  The  period  lasts  for  80  to  100  to  160  to  180 
years  and  also  is  subdivided  into  two  phases.  During  the 
first  phase  (3),  the  upper  story,  which  is  dominated  by 
dark-taiga  coniferous  species  is  formed.  The  second  phase 
(4)  is  characterized  by  stabilization  of  economically  ma- 
ture spruce  and  fir  forest,  high  seed  production  of  all 
forest-forming  species,  initial  mortality  first  of  fir  and 
then  of  spruce,  and  then  the  appearance  of  second-growth 
generations  of  dark-taiga  coniferous  trees  and  stone  pine. 

Third  Period — Stabilization  and  natursd  maturity 
of  stone  pine  forests.  This  period  begins  after  160  to  180 
years  and  persists  400  years  and  longer.  Stone  pine  is 
dominant,  its  share  in  the  canopy  layer  amounts  to  30  to 
100  percent.  First  the  lower  layer  of  woody  plants  and 


then  the  medium  layer  comprising  second-growth  genera- 
tions are  formed  in  this  period.  The  period  can  be  divided 
into  two  phases:  In  the  first  phase  (5),  stone  pine  forests 
are  stabilized,  and  the  increment  of  merchantable  timber 
is  at  its  maximum.  Stone  pine  is  producing  maximum 
amounts  of  seed,  and  the  mortality  of  old  fir  gind  spruce 
trees  continues.  The  second  phase  (6)  is  characterized  by 
maturity  of  stone  pine  forests,  decrease  in  seed  produc- 
tion, gradual  degradation  of  the  older  (first)  canopy  layer, 
and  then  the  formation  of  the  main  story  by  younger  gen- 
erations of  dark-taiga  coniferous  trees  and  stone  pines. 
Phase  duration  varies  between  280  to  300  to  400  to  450 
years,  and  more. 

Sometimes  a  third  phase  becomes  apparent,  when  a 
few  old,  large-sized  stone  pines  dominate  the  canopy 
layer,  while  the  main  forest  story  is  formed  by  second  and 
younger  generations  of  deciduous  trees,  spruce,  fir,  and 
stone  pine.  The  proportion  of  stone  pine  in  forest  compo- 
sition is  different  in  various  regions  and  forest  sites.  In 
pretimdra  forests,  in  subzones  of  northern  and  partially 
middle  taiga  of  western  Siberia,  in  mountain  stone  pine 
forests  of  the  Ural  Mountains,  and  on  fresh  sites,  stone 
pine  is  very  common  in  second-generation  stands.  Under 
humid  conditions  in  the  moist  sites  of  the  middle  and 
southern  tedga,  and  in  the  foothills  and  low-elevation 
areas  of  the  Altai-Sayan  mountain  system,  stone  pine 
forests  are  often  replaced  by  spruce  and  fir  forests. 


80 


c: 
a> 

o  60 

O) 
Q. 

o 
E 

3 

o  40H 
> 


20 


Period  1 


Period  2 


Period  3 


Siberian  Stone  Pine 
Siberian  Larch 


60 


80  120  160  200 

Age  (years) 


240 


280 


Figure  3 — Reproduction-age  dynamics  scheme  of  reedgrass-vaccinium, 
stone  pine  forests  on  the  northern  macroslope  of  eastern  Tannu-Ola 
(Tuva-northern  Mongolia).  Stand  admixture  variation:  1— Siberian  stone 
pine;  2 — Siberian  larch. 


278 


The  regularities  described  here  in  reproduction-age 
dynamics  are  surely  far  from  reflecting  the  whole  variety 
observed  in  nature,  but  they  are  clesirly  determined  by 
ecological-geographical  factors.  For  example,  figure  3 
illustrates  reproduction-age  dynamics  of  a  stone  pine  for- 
est with  additional  larch,  located  on  the  northern  slope 
of  eastern  Tannu-Ola  (at  the  Russian-Mongolian  border). 
However,  despite  all  this  there  also  are  general  regulari- 
ties, which  are  presented  in  the  generalized  scheme. 
They  are  due  to  the  most  essential  ecological-biological 
properties  of  stone  pine,  the  major  forest-forming  species. 
These  stone  pine  characteristics  include  seed  dispersal  by 
animals  (strongest  relationship  with  Nucifraga  caryocat- 
actes  macrorhynchos  Br.),  relative  shade  tolerance  at 
early  stages  of  existence,  accelerated  growth  and  develop- 
ment under  favorable  light  conditions,  and  a  much  longer 
life  span  than  other  forest-forming  species. 

FOREST  MANAGEMENT 
IMPLICATIONS 

An  analysis  of  forest  management  materials  shows  that 
the  regularities  in  the  reproduction-age  dynamics  are  not 
adequately  considered  in  forest  management.  Stands  of 
the  initial  period  are  usually  taken  for  deciduous  forests, 
and  those  of  the  second  period  as  spruce  and  fir  forests, 
with  management  norms  assigned  correspondingly.  This 
leads  inevitably  to  an  artificial  alternation  of  tree  species, 
to  cutting  of  actually  young  and  middle-aged  stone  pine 
forests.  All  the  investigation  data  suggest  the  need  for  an 
integrated  stone  pine  forestry  program,  which  would  com- 
bine potential  stone  pine-deciduous  forests  of  the  first  pe- 
riod, spruce  and  fir  forests  of  the  second  period,  and  stone 
pine  forests  of  the  third  period.  Thus,  a  strategic  basis  for 
the  development  of  a  management  system  could  be  created. 


If  this  basis  is  not  available,  the  areas  covered  with  stone 
pine  forests  will  surely  continue  to  shrink. 

REFERENCES 

Bobrov,  E.  G.  1974.  Coniferous  breeds  of  USSR.  Nauka. 
212  p.  [In  Russian]. 

Clements,  F.  E.  1928.  Plant  succession  and  indicators. 
New  York:  Hafner  Publishing.  453  p. 

Kolesnikov,  B.  P.  1956.  Stone  pine  of  forests  in  Far  East. 
Publ.  Academy  of  Sciences  of  USSR.  264  p.  [In  Russian]. 

Kolesnikov,  B.  P.;  Zubareva,  R.  S.;  Smolonogov,  E.  P. 
1974.  Forest-site  conditions  and  forests  types  in 
Sverdlovsk's  region.  Publ.  Ural  Division  of  Academy  of 
Sciences  USSR.  Sverdlovsk.  176  p.  [In  Russian]. 

Krylov,  G.  V.  1961.  Forests  in  west  Siberia.  Publ.  Acad- 
emy of  Science  of  USSR.  255  p.  [In  Russian]. 

Luzganov,  A.  G.;  Abaimov,  A.  P.  1977.  Role  of  river  basins 
and  wind  in  spreading  and  evolution  of  larch,  stone  pine 
and  others  forest  breeds.  Publ.  Siberian  Technological 
Institute.  8:  31-38.  [In  Russian]. 

Neishtadt,  M.  I.  1957.  History  of  forests  and  paleogeogra- 
phy  of  USSR  in  holocene.  Publ.  Academy  of  Sciences  of 
USSR.  404  p.  [In  Russian]. 

Nepomilueva,  N.  I.  1974.  Siberian  stone  pine  (Pinus 
sibirica  Do  Tour.)  in  north-east  of  European  part  of 
USSR.  Nauka.  184  p.  [In  Russian]. 

Smolonogov,  E.  P.  1990.  Ecological  and  geographical  dif- 
ferentiation and  dynamics  of  Siberian  stone  pine  forests 
in  Ural  and  plains  of  West  Siberia.  Publ.  Ural  Division 
of  Academy  of  Sciences  of  USSR.  Svezdlovsk.  286  p.  [In 
Russian]. 

Smolonogov,  E.  P.;  Vegerin,  A.  M.;  Kolesnikov,  B.  P. 
1970. Forest-site  zones  in  Tumen's  region — botanical 
researches  in  Ural.  Publ.  Ural  Division  of  Academy  of 
Sciences  of  USSR.  Sverdlovsk:  34-58.  [In  Russian]. 


279 


THE  CEMBRAN  PINE  IN  THE  FRENCH 
ALPS:  STAND  DYNAMICS  OF  A 
CEMBRAN  PINE  FOREST  IN  TUEDA 
(SAVOY,  FRANCE) 

Lise  Wlerick 


Abstract— In  the  Tueda  Natural  Reserve  (Savoy,  France),  the 
most  important  northern  French  cembran  pine  stand,  a  study  of 
stand  structure  dynamics  was  started  in  mixed  stands  of  spruce 
(Picea  abies)  and  cembran  pine  (Pinus  cembra).  Six  types  of  stand 
structures  were  estabhshed  on  the  basis  of  the  following  main  cri- 
teria: spruce  and  cembran  pine  ratio,  extent  of  canopies,  and  re- 
generation and  diameter  classes.  The  stands  are  very  young. 


History 

Where  cembran  pine  grew  in  the  upper  part  of  the  forests, 
it  was  often  eliminated  to  create  meadows  and  allow  graz- 
ing. Larch  {Larix  decidua  Miller),  on  the  other  hand,  with 
its  less  dense  canopy  could  shelter  animals  as  well  as  let 
them  graze  the  grass  under  it.  Despite  the  fact  that  larch 
grows  at  the  same  altitude  as  cembran  pine,  it  suffered 
less  from  human  impact  than  the  latter. 

Upper  montane  and  subalpine  belt  forests  were  subjected 
to  fire  and  excessive  deforestation  and  logging.  Much  wood 
was  used  for  cooperage  (wooden  casks),  millwork,  cabinet- 
work, and  wood  carving  (Contini  and  Lavarelo  1982), 

The  decrease  of  cembran  pine  is  mainly  due  to  human 
influence.  Palynological  studies  in  French  alpine  peatlands 
reveal  old  cembran  pine  occurrences  outside  its  present 
range.  Fourchy  (1986)  cited  the  cembran  pine  occurrence 
of  FeUeautier  west  of  Gap;  Bartoli  (1966)  mentioned  an 
occurrence  in  the  peatlands  in  the  Valmeinier  Valley  in 
High  Maurienne  as  well  as  another  one  in  areas  above 
Thyl,  which  are  treeless  now. 

Within  its  present  distribution  area,  cembran  pine 
mainly  occurs  on  northern  and  steep  slopes.  On  the  south- 
em  slopes,  only  a  few  cembran  pines  are  left,  since  those 
areas  were  traditionally  preferred  to  build  houses  and  for 
farming  and  grazing,  just  as  on  gentle  slopes.  The  Versant 
du  Soleil  above  La  Cote  d'Aime  in  Tarentaise  is  an  excel- 
lent example  of  this  use  of  space. 

The  distribution  of  cembran  pine  must  have  been  at  its 
maximum  3,000  or  4,000  years  ago,  before  humans  started 
substantial  deforestation  (Combes  1986).  Moimtfdn  popu- 
lations began  to  decrease  after  1850  (Ozenda  1985).  The 
consequences  of  this  change  on  logging,  fire,  and  pasture 
became  obvious  between  1860  and  1880,  according  to  de- 
partments. An  increase  of  cembran  pine  in  subalpine  heaths 
and  in  areas  no  longer  used  for  grazing  has  been  observed 
for  more  than  40  years.  At  present,  we  observe  a  great  vital- 
ity of  cembran  pine,  which  is  spreading  horizontally  and  to 
lower  elevations  (Maurienne:  1,100  m;  Tarentaise:  1,300  m), 
often  far  from  seed-bearing  trees.  It  does  not  extend  much 
to  higher  elevations. 

Natural  Environment 

Cembran  pine  extends  from  the  Alpes  Maritimes  to 
the  Haute  Savoie.  Its  distribution  is  roughly  the  same  as 
that  of  larch  (fig.  1),  except  for  two  main  differences:  in  the 
northern  Alps  its  western  boundary  greatly  overlaps  the 


Cembran  pine  (Pinus  cembra  L.)  has  been  heavily  over- 
exploited  in  France.  Nowadays  it  can  only  be  seen  in  mar- 
ginal areas  with  an  often-limited  accessibility.  Queyras 
(Hautes  Alpes)  and  Maurienne  (Savoy),  however,  still  pos- 
sess beautiful  cembran  pine  stands.  In  the  north,  cembran 
pine  is  often  scattered.  Nevertheless,  in  the  Tueda  forest 
(in  the  Allues  Parish)  the  most  remarkable  cembran  pine 
stand  exists  in  the  northern  French  Alps. 

Its  importance  and  specificity  led  us  to  study  it  more 
closely.  The  possibility  of  a  future  intervention  encour- 
aged us  to  learn  about  its  history,  its  characteristics,  and 
its  stand  dynamics. 

FRENCH  CEMBRAN  PINE  STANDS 

Cembran  pine  is  present  throughout  the  Alps  from  the 
Mediterranean  Sea  to  eastern  Austria  (Contini  and  Lavarelo 
1982).  It  grows  mainly  in  the  Alpine  Region  but  can  also 
be  found  in  the  Carpathian  Mountains  and  the  Transyl- 
vanian  Alps.  In  the  Alps,  cembran  pine  can  only  be  found 
in  Austria,  France,  Italy,  and  Switzerland.  The  amount 
of  cembran  pine  stands  varies  from  one  country  to  another, 
and  occurs  in  the  following  decreasing  order: 

•  19,400  ha  in  Italy  (Del  Favero  and  others  1985). 

•  15,600  ha  in  Austria  (Contini  and  Lavarelo  1982). 

•  12,400  ha  in  France  (Contini  and  Lavarelo  1982). 

•  11,900  ha  in  Switzerland  (National  Forestry  Inventory). 

These  figures  come  from  the  National  Forestry  Inventories. 

The  distribution  of  cembran  pine  is  mostly  contiguous 
in  Austria,  Italy,  and  Switzerland;  in  France  it  is  sporadic. 
This  is  essentially  due  to  strong  human  influences. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11, 1992. 

Lise  W16rick  is  a  Forester,  National  Forestry  Office,  Technical  Division, 
73200  Albertville.  France. 


280 


larch  distribution  while,  in  the  southern  Alps,  larch  goes 
farther  south  than  cembran  pine  (Fourchy  1986).  However, 
Combes  (1986)  showed  in  the  Alpes  of  Haute  Provence  the 
cembran  pine  spread  farther  south  than  Fourchy  found  it. 

We  can  divide  Alps  into  three  parts:  Pre-Alps,  Intermedi- 
ate Alps,  and  Intra-Alpine  Alps.  Pre-Alps  can  be  character- 
ized by  the  widespread  natural  distribution  of  beech  {Fagus 
sylvatica  L.)  and  by  the  absence  of  larch  growing  naturally; 
the  Intra-Alpine  axis  by  natural  occurrence  of  larch  and 
the  absence  of  beech;  and  the  Intermediate  Alps  by  the  co- 
existence of  natiu-al  larch  with  natural  beech.  According  to 
these  definitions,  in  the  northern  Alps  cembran  pine  reaches 
the  Pre-Alps  with  cembran  pine  stands  spread  out  fi*om 


0     «  so  «0  Km 

 natural  distribution  of  larch 

 natural  distribution  of  cembran  pine 


Figure  1 — Distribution  of  cembran  pine  and  larch  in 
the  French  Alps  (Fourchy  1968).  1— Haute  Savoie; 
2 — Savoie;  3 — Is6re;  4 — Hautes  Alpes;  5 — Alpes  de 
Haute  Provense;  6 — Alpes  maritimes.  According  to 
my  personal  observation,  the  boundary  of  larch  has 
been  modified  in  Tarentaise  (  )  just  like  that  of 
cembran  pine  in  Alpes  de  Haute  Provence 
as  mentioned  in  Combes  (1986)  (  ). 


1,500  to  2,000  m  of  altitude,  while  isolated  or  stxinted  trees 
occur  fi"om  1,300  to  2,000  m  of  altitude.  In  the  southern 
Alps,  cembran  pine  is  limited  to  Intra-Alpine  and  Interme- 
diate Alps;  stands  are  spread  out  from  1,800  to  2,200  m  of 
altitude,  stimted  or  isolated  trees  from  1,100  to  2,480  m.  A 
recolonization  of  this  species  particularly  at  lower  altitudes 
has  been  observed  in  recent  years  (Tarentaise,  Maurienne). 

All  exposures  are  fit,  but  a  clear  preference  for  cool  expo- 
svu*es  (northern  and  neighboring)  is  known.  It  is  not  obvious 
whether  this  distribution  is  due  to  ecology  or  to  human  in- 
fluence (more  important  in  simny  exposures)  (Foiirchy  1968). 

Cembran  pine  can  grow  on  any  geological  substrata  and 
more  favorably  on  sihceous  rocks:  gneiss,  granite,  sandstone, 
flysch,  schist,  etc.  It  can,  however,  be  found  on  gypsvmi  in 
La  Plagne  (Tarentaise)  or  limestone  in  Flaine  (Fourchy 
1968)  and  on  gypsum  and  limestone  in  Termignon  and 
Bramans  (Maurieime)  (Bartoli  1966).  Sites  on  limestone 
are  usually  rather  marginal. 

We  can  differentiate  hygrophilous  subseries  from  xero- 
philous  subseries  within  communities.  The  hygrophilous 
subseries  characterized  by  Pinus  cembra,  Rhododendron 
ferrugineum,  Lonicera  caerulea,  Vaccinium  myrtillus,  Vac- 
cinium  vitis-idaea,  Calamagrostis  villosa,  etc.,  which  can 
be  separated  into  four  facies:  Vaccinium  myrtillus  fades. 
Geranium  sylvaticum  facies  in  coomb,  Calamagrostis  villosa 
facies,  and  Pinus  uncinata  facies,  with  Erica  herbacea, 
Sorbus  chamaemespilus,  etc.,  on  Limestone.  The  xerophilous 
subseries  differs  in  the  presence  of  Juniperus  nana  and  Arc- 
tostaphylos  uva-ursi. 

Along  the  subalpine  belt  two  ecological  situations  can 
be  observed,  one  on  the  northern  slope  and  one  on  the 
southern  slope.  These  coincide  with  two  heath  types,  one 
of  Rhododendron-Vaccinium  and  the  other  of  Juniperus- 
Arctostaphylos,  respectively,  and  are  liable  to  afforest  in 
one  or  the  other  of  the  four  tree  species:  cembran  pine, 
larch,  moimtain  pine  (Pinus  uncinata  [DC]),  and  spruce 
(Picea  abies  [L.])  (Ozenda  1985). 

In  the  northern  Alps  the  former  situation  prevails.  The 
climate  is  wetter,  and  as  a  resiilt  the  difference  in  exposure 
between  the  south-facing  slope  and  the  north-facing  slope 
is  less  perceptible  than  in  the  southern  Alps — moisture  is 
rarely  a  limiting  factor.  Therefore,  cembran  pine  stands, 
or  cembran  pine  stands  with  larch,  represent  the  most  pre- 
vailing types  and  that  xerophilous  type  tends  to  be  limited 
to  the  southern  slopes. 

Structure  and  Dendrometrics 

In  France,  it  is  uncommon  to  see  pure  cembran  pine 
stands,  such  as  those  seen  in  the  Engadine  or  Tyrol 
(Fourchy  1968).  We  find  this  pine  often  associated  with 
spruce  or  mountain  pine  in  Pre-Alps  and  with  larch  or 
mountain  pine  in  the  Intra-Alpine  axis.  French  National 
Forestry  Inventory  reports,  in  decreasing  order,  the  associ- 
ate species  of  cembran  pine  (Contini  and  Lavarelo  1982): 


Species  Percent 

Larch  86.0 

Spruce  30.0 

Mountain  pine  15.0 

Scots  pine  10.0 

Fir  9.0 
Various  broad-leaved  trees 

(aspen,  alder,  birch)  4.5 


281 


Cembran  pine  stands  usually  have  sparse  canopies. 
Stand  density  is  often  less  than  200  trees/ha  with  a  me£in 
at  the  same  elevation  of  135  trees/ha  (Contini  and  Lavarelo 
1982).  Other  species  have  higher  stand  densities.  Standing 
volume  per  hectare  is  also  low  (35  m^/ha  on  the  average). 

Structure  of  natviral  stands  is  always  an  irregrilar  clump 
forest.  It  is  interesting  to  note  that  even  in  the  case  of  high 
stand  density,  we  have  the  same  type  of  structure:  clumps 
separated  by  small  openings  whose  width  is  about  one  to 
three  times  as  much  as  the  average  diameter  of  the  crown 
of  the  biggest  trees  (Contini  and  Lavarelo  1982). 

Cembran  pine  very  easily  colonizes  lands  abandoned  by 
grazing  and  human  activity;  it  also  regenerates  well  under 
clear  stands.  On  the  other  hand,  it  does  not  regenerate 
easily  under  denser  crown  cover. 

DYNAMICS  OF  THE  TUEDA  STAND 

The  cembran  pine  stand  of  Tueda  constitutes  a  remark- 
able upper  altitude  ecosystem,  with  respect  to  ecology  and 
landscape,  on  the  outskirts  of  Vanoise  National  Park  in  the 
Allues  Parish  (also  see  fig.  1  for  its  location).  It  has  been 
managed  by  the  National  Forestry  Office  since  1989,  and 
jointly  with  the  National  Park  of  Vanoise  since  1990,  when 
the  Natural  Reserve  of  Tueda,  which  includes  the  entire 
stand,  was  established.  The  National  Forestry  Office  man- 
ages about  600  ha  out  of  the  1,100  ha  of  the  natural  reserve. 

Of  the  600  hectares  there  are  150  ha  of  more  or  less 
dense  pure  cembran  pine  stands  in  the  upper  part  and 
mixed  cembran  pine  and  spruce  stands  in  the  lower  part 
(Pignol  1992).  These  stands  range  from  1,670  to  2,220  m  in 
altitude.  Above  1,900  to  2,000  m  altitude,  spruce  is  rarer 
and  varies  by  exposure.  The  forest  is  essentially  rooted  in 
acid  rocks  (schist,  sandstone,  gneiss)  (Pignol  1992).  This 
area,  which  has  been  grazed  for  a  long  time  by  cows  and 
goats,  is  still  grazed  today,  though  less  extensively  than 
before. 

We  estimate  that  the  cembran  pine  stand  will  eventually 
extend  to  about  400  ha.  That  is  400  ha  of  land  which  will 
become  forest  depending  on  how  rapidly  grazing  disappears 
and  if  humans  do  not  intervene. 

Patches,  which  were  private  until  1969,  have  not  been 
logged  since  that  time.  However,  when  the  patches  were 
bought  by  the  village  council,  a  lot  of  large-size  trees  had 
already  been  logged  by  the  owners.  Apart  from  a  few  wind- 
throws,  the  forest,  which  is  young,  is  getting  older.  As 
meadows  and  pastures  become  afforested,  cembran  pine 
and  spruce  seedlings  quickly  colonize  open  spaces,  and 
the  pressure  of  grazing  is  not  strong  enough  to  fight  this 
phenomenon. 

At  present  there  is  an  interesting  balance  in  the  land- 
scape between  the  more  or  less  clear  stands,  the  pastures, 
the  meadows,  the  Alnus  viridis  bushes,  Rhododendron  and 
Juniperus  heaths,  and  fallen  rocks. 

Study  Purposes 

The  cembran  pine  stand  of  Tueda  is  the  most  important 
northern  cembran  pine  stand  in  France.  What  is  interest- 
ing is  that  it  is  a  young  stand.  Recent  pollen  analyses  have 
confirmed  the  absence  of  larch  and  mountain  pine.  The  as- 
sociation of  cembran  pine  with  spruce  to  such  a  large  extent 


is  unique  in  France  and  very  rare  in  Europe  (a  few  stands 
in  Italy  and  Austria). 

Besides  being  a  very  rare  occurrence  in  the  French  Alps, 
the  relict  cembran  pine  stand  of  Tueda  also  offers  the  ad- 
vantage of  presenting  side  by  side  two  t)T)ical  associations 
of  cembran  pine  on  three  different  exposures  (southwest, 
north,  northeast): 

•  the  hygrophylous  subseries  with  Rhododendron 
ferrugineum  and  Vaccinium  myrtillus. 

•  the  xerophylous  subseries  with  Juniperus  nana. 

Besides  the  usual  varied  mountain  flora,  the  cembran  pine 
stfind  of  Tueda  contains  the  most  luxuriant  Linnaea  borealis 
site  of  the  six  French  sites,  all  found  in  Savoy.  Linnaea  bo- 
realis is  a  true  boreal  pine  relict  protected  at  the  national 
level.  We  also  find  a  lot  of  wet  sites  with  peatlands  and  pro- 
tected flowers  such  as  Stemmacantha  rhapontica,  Clematis 
alpina,  Primula  farinosa,  Swertia  perennis,  Carex  micro- 
glochin,  and  Carex  bicolor. 

Wildlife  is  just  as  rich.  We  observe  nutcrackers,  black 
grouse,  golden  eagles  (which  nest  there),  rock  partridge, 
and  ptsirmigan,  as  well  as  many  other  species. 

To  manage  this  natural  area  with  the  aim  of  preserving 
or  improving  its  biodiversity  and  guaranteeing  an  always 
attractive  landscape  for  the  many  visitors  who  walk  throiigh 
it  in  summer  and  in  winter,  we  wish  to  know  better  the 
stand  dynamics.  We  thus  are  studying  the  natural  growth 
of  the  existing  stands  as  well  as  the  competition  between 
cembran  pine  and  spruce.  We  also  are  studying  the  mecha- 
nisms of  the  closing  of  present  openings  and  the  rising  of 
the  timberline,  which  seem  to  be  linked  to  the  presence 
of  the  nutcracker,  a  bird  that  favors  the  scattering  of  seeds. 
We  will  obtain  information  that  will  allow  us  to  decide 
whether  it  is  necessary  to  intervene  or  not  in  this  sensitive 
natural  area,  and  if  we  do,  how. 

Cembran  pine,  a  noble  coniferous  species  used  for  wood- 
carving  and  cabinetwork,  has  been  heavily  overexploited 
up  to  now.  This  is  a  much-prized  species  from  an  economic 
and  cultural  point  of  view.  Its  price  is  five  times  as  high  as 
that  of  spruce. 

Many  forests  in  Tarentaise  and  in  Mamienne,  as  well  as 
in  the  rest  of  the  Alps,  are  likely  to  be  favorable  for  the  natu- 
ral development  or  forestry  planting  of  cembran  pines.  The 
study  of  the  cembran  pine  stand  of  Tueda  will  also  be  of 
great  use  for  other  cembran  pine  stands  (present  or  potential) 
in  Savoy,  both  in  the  French  Alps  and  the  rest  of  Europe. 

Methodology 

A  forest  management  plan  was  established  in  1991.  On 
that  occasion  the  vegetation,  stand  and  regeneration  map- 
ping was  carried  out. 

Four  types  of  stands  were  distinguished: 

•  pure  cembran  pine  stands  with  5  to  40  percent  canopy 
and  with  >40  percent  canopy. 

•  mixed  spruce  and  cembran  pine  stands  with  5  to  40 
percent  canopy  and  with  >40  percent  canopy. 

For  regeneration,  the  presence  of  cembran  pine  and  spruce 
seedlings  was  noted  both  under  crown  cover  or  beside  it. 

These  mappings  allowed  us  to  develop  a  statistical  in- 
ventory in  the  mixed  spruce  and  cembran  pine  stands  in 


282 


1992.  One  plot  with  5  are/ha  (1  are  =  100  m^)  was  counted 
in  a  systematic  grid.  In  total  75  plots  were  computed.  The 
following  data  and  measurements  were  taken: 

•  environment  data:  altitude,  exposure,  main  vegetation, 
presence  or  absence  of  falling  rocks. 

•  dendrometrical  data:  (1)  inventory  of  all  trees  over 
7.5  cm  diameter  by  species  and  by  diameter  classes;  (2)  sam- 
ple basal  area;  (3)  on  the  three  trees  nearest  to  the  center 
of  the  plot:  width  of  the  last  10  rings,  age  at  1.30  m,  diam- 
eters of  crown  projection  (maximum  diameter  and  diameter 
perpendicular  to  the  maximmn  diameter),  total  height, 
height  to  first  live  branch. 

•  morphological  data:  stand  structure  over  25  ares  (total 
height  was  considered  as  absolute  height). 

The  maximum  height  was  estimated  to  be  24  m.  Thus 
the  creation  of  four  height  strata: 


Height  (m) 
24 

18 

12 

6 

0 


Strata 

Stratum  1 

Stratum  2 

Stratiun  3 

Stratum  4= 
regeneration 


The  ratios  of  strata  1,  2,  and  3  have  been  noted  in  1.10~\ 
Three  categories  of  diameter  were  defined:  small-size 
trees  (PB)  10  cm  <  0  <  20  cm;  medium-size  trees  (BM) 
25  cm  <  (j)  <  25  cm;  large-size  trees  (GB)  <J)  >40  cm. 

The  ratios  of  these  three  categories  of  diameter  have  also 
been  given: 

•  Presence  of  openings  covering  a  surface  of  at  least  10 
ares  out  of  25  ares  have  been  noted  in  1.10'\ 

•  Presence  and  abundance  of  regeneration  over  5  ares 
have  been  classified  in  four  groups: 


Group 

0 
1 
2 
3 


Regeneration 

none 

very  diffuse 

scattered 

abundant 


The  ratio  between  spruce  and  cembran  pine  regeneration 
also  has  been  noted  in  1.10''. 

•  Windthrow  number  over  5  ares  for  spruce  and  cembran 
pine. 

•  Stump  number  over  5  ares. 

•  Canopy  (crown  projection  of  strata  1,2,  and  3  in  ratio 
at  the  total  surface  of  the  plot)  over  5  ares  in  1.10"'. 

Data  Processing  and  Results 

A  first  principal  component  analysis  has  enabled  us  to 
leave  out  the  nonsignificant  variables  and  those  that  are 
too  correlated.  Sixteen  variables  out  of  29  have  thus  been 
kept  for  the  second  principal  component  analysis.  They  can 
be  foimd  in  figure  4.  The  first  three  main  axes  of  the  second 


principal  component  analysis  account  for  61.3  percent  of  the 
inertia  of  the  phenomenon.  The  first  main  axis  accounts  for 
29.6  percent  of  the  total  variation,  the  second  one  for  17.8 
percent,  and  the  third  one  for  13.9  percent.  These  results 
are  satisfactory  (Leclerc  1992). 

A  hierarchical  ascending  classification  then  enabled  us 
to  determine  six  different  types  of  stands  and  to  draw  up 
a  determination  key. 

Group  N°l  is  characterized  by  a  very  high  percentage  of 
cembran  pine  (91  percent  of  the  number  of  trees),  a  medium 
canopy  (5.4*10"^),  very  low  regeneration,  very  few  small- 
size  trees  (0.8*10"^),  quite  a  few  large-size  trees  (5.3*10"^), 
and  an  average  nimiber  of  medium-size  trees  (3.9*10"^).  It 
can  be  assimilated  to  an  irregular  cembran  pine  clump  for- 
est, with  large-  and  mediimi-size  trees  and  of  average  stand 
density,  in  which  large-size  trees  clearly  predominate. 

Group  N°2  is  characterized  by  a  higher  proportion  of 
spruce,  which  however  remains  moderate  (55  percent  of 
the  number  of  trees),  an  important  canopy  (7.7*10"'),  an 
average  rate  of  regeneration  in  which  spruce  predominates, 
quite  a  balanced  distribution  of  small-size,  medium-size,  and 
large-size  trees  (respectively,  4.0*10"^;  3.6*10"';  2.4*10"').  It 
looks  like  a  selection  forest  of  cembran  pine  and  spruce  in 
which  spruce  predominates. 

Group  N°3  has  a  very  high  percentage  of  cembran  pine 
(84  percent  of  the  number  of  trees),  but  it  differs  from 
Group  N°l  by  a  less  dense  canopy  (3.6*10"'),  regeneration 
exclusively  composed  of  rather  scattered  cembran  pine  seed- 
lings, a  hardly  marked  predominance  of  medium-size  trees 
(4.5*10"'),  a  not  inconsiderable  number  of  large-size  trees 
(3.9*10"'),  and  few  small-size  trees  (1.6*10"').  It  is  quite 
similar  to  an  irregular  cembran  pine  clump  forest  with  me- 
dium and  large  trees,  of  low  stand  density,  but  with  a  pre- 
dominance of  medium-size  trees  and  rather  diffused  regen- 
eration of  cembran  pine. 

Group  N°4  is  also  characterized  by  a  high  percentage  of 
cembran  pine  (7.4*10"'  of  the  number  of  trees),  a  limited 
canopy  (3.5*10"'),  rather  diffused  regeneration  of  spruce  and 
cembran  pine,  a  predominance  of  small-size  trees  (6.0*10"'), 
a  limited  proportion  of  medium-size  trees  (2.9*10"'),  and 
few  large-size  trees  (1.1*10"').  It  looks  like  a  mixed  spruce 
and  cembran  pine  clvimp  forest,  but  with  a  predominance 
of  cembran  pine,  of  low  stand  density,  with  a  diffused  regen- 
eration of  spruce  and  cembrsin  pine,  rather  regular  in  its 
small-size  trees.  Altitude  is  between  1,680  and  1,850  m. 

Group  N°5  differs  from  Group  N°4  in  that  it  is  located 
at  a  higher  altitude,  from  1,780  to  1,950  m.  Moreover,  the 
regeneration  is  abundant,  with  a  predominance  of  cembran 
pine.  The  canopy  is  of  medium  density  (4.5*10"'),  the  pro- 
portion of  cembran  pine  is  high  (77  percent  of  the  number 
of  trees),  and  the  small-size  trees  largely  predominate 
(7.1*10"').  It  can  be  assimilated  into  a  regular  cembran  pine 
forest  of  small-size  trees,  with  medium  stand  density  and 
abundant  regeneration. 

Group  N°6  is  very  different  from  the  other  groups.  The 
proportion  of  spruce  is  very  important  (86  percent  of  the 
number  of  trees).  So  are  the  canopy  (9.5*10"')  and  the  pro- 
portion of  small-size  trees  (7.8*10"').  However,  these  small- 
size  trees  are  distributed  over  two  height  strata,  stratvun  3 
(6  to  12  m)  and  stratum  2  (12  to  18  m).  Regeneration  is 
of  course  limited  for  this  group,  which  is  similar  to  a  very 
dense  spruce  stand,  with  very  regular  small-size  trees. 


283 


To  illustrate  all  this,  the  characteristics  of  groups  1  and  The  factorial  discriminant  analysis  has  enabled  us  to 

6  are  given  in  figures  2  and  3.  bring  out  the  correlations  between  the  variables  (fig.  4). 


1 

1  Observations 

;           — — _ — ^ — , — . 

1       Spruce  and  cembran  pine 

1 

1 
1 

Spruce  (sp) 

Cembran  pine  (cp) 

1 

1  Average  altitude  (m)  -  1 .740 

1 
1 

Spruce  regeneration 

Cembran  pine  regeneration  1 

1 

1 

(1/10) 

0 

(1/10) 

1.4  1 

1    g  calculated 

1  Canopy                 (1/10)  - 

5.4  1 

1  Stratum  1  (24-18  m)   (1/10)  = 

1.9  1 

N  sp/ha 

28 

Ncp/ha 

274 

1  Stratum  2  (18-12  m)   (1/10)  - 

5.4  1 

g  sp  (m*)  * 

3 

I  gcp(m'0* 

36.4  1 

1  Stratum  3  (12-6  m)     (1/10)  - 

2.7  1 

V  sp  (m3/ha) 

22.5 

Vcp(m3/ha) 

205.3  1 

1  GB  -  0  ^  40  cm 

1  GB                       (1/10)  ° 

5.3  j 

GBsp(1/10) 

4 

GBcp(1/10) 

4.4  1 

1  BM  -  25  cm  <  0  <  35  cm 

1  BM                       (1/10)  - 

3.9  1 

BMsp(1/10) 

2 

BMcp(1/10) 

4.3  1 

I  PB=  10cmS(t)<20cm 

1  PB                       (1/10)  - 

0.8  1 

PBsp(1/10) 

4 

PBcp(1/10) 

1.3  1 

1       Regeneration  code 

1  Regeneration  (code) 

0.1  1 

%  sp  in  number  of 

%  cp  in  number  of 

1  0  -  no  regeneration 

1  Total  number  of  trees/ha 

302  1 

trees 

9 

trees 

91  j 

1  1  -  very  dittuse 

1  g  (m^)  sample  basal  area 

17.2  1 

%  sp  in  g 

8 

1  2  =  scattered 

1  3  =  abundant 
1 

1 

1 

1 

m 


1234526         4  2  3 


1  -  GB  -  ST  2  -  sp  4  »  GB  -  ST  1  -  cp 

2  =  BM  -  ST  2  -  cp  5  =  PB  -  ST  3  -  cp 

3  =  BM  -  ST  3  -  cp  6  =  regeneration  -  cp 

Figure  2 — Group  1 :  Characteristics  and  diagram. 


284 


1  Observations 

1           Cnrii/^a  ttnH  /^AmhrAH  olnA 
1          oprucv  anu  voiiiuiciii  pmo 

Soruce (sp) 

"""^ 

Cembran  pine  (cp) 

1 

 1 

1  AwArana  sltitiiHA  lm\  —  1 
1  MVoiay"  dUllUUc  V'"/  ~  1 .0  I 

Spruce  regeneration 

Cembran  pine  regeneration  1 

(1/10) 

4.5 

(1/10) 

0.5  [ 

1    g  calculated 

1       nr\r\r\\/                                 ( A  /  ^C^\  s 

1  oanopy                 \*'  ^ 

9.5  1 

Qtratiim  1  f^A-AR  m\  « 
1  OualUili  1          lo  III/     V  ' 

1  j 

N  sp/ha              -  1 ,880 

N  cp/ha 

310 

1  Qtratiim  9  Mft-19  m\  s 

4  1 

g  sp  (m^^ 

56.5  ; 

g  cp  (m^  * 

9  1 

1  Qtratiim     M9.fi             M/IO^  s 

1  oiraium  o  ^  it-o  wi)      \  1/  ivj/  *= 

5  1 

V  sp  (m3/ha) 

328.3 

V  cp  (m3/ha) 

33.2  1 

1    OD  *  y  WTl 

1  GB                           (1/101  « 

0.4] 

GBsp(1/10) 

0.4 

GBcp(1/10) 

0.6  1 

1    DiUi  _  QC  rrn  <  A  < 

1    DiVI  ■  ^9  uiM  ^  vp  ^  uiii 

i  BM                         (1/10)  - 

i.sl 

BMsp(1/10) 

1.9 

BMcp(1/10) 

1.0  j 

1    PR  «  1  n  rm  <  A  <  20  cm 

1  PB                          (1/10)  - 

7.8 1 

PBsp(1/10) 

7.7 

PBcp(1/10) 

8.4  1 

1         RAnArtArAtinn  cods 

1  Regeneration  (code)  » 

0.5 1 

%  sp  in  number  of 

%  cp  in  number  of 

1  0  «  no  regeneration 

1  Total  number  of  trees/ha 

2,190  1 

trees 

86 

trees 

14  j 

1  1  =  very  diffuse 

1  g  (m^)  sample  basal  area 

34  1 

%  sp  in  g 

86 

1  2  «  scattered 

1  3  =  abundant 

1 

1 

1 

1 

2 

2  8 

3     15  112 

4 

1 

1       2  1 

2     116    2  1 

1  = 

PB 

-ST  3 

-sp 

5  =  BM  -  ST  2  -  sp 

2  = 

PB 

-ST  2 

■sp 

6  =  GB  -  ST  1  -  cp 

3  = 

PB 

-ST  3 

■cp 

7  =  BM  -  ST  2  -  cp 

4  = 

GB 

ST  1  - 

cp 

8  =  regeneration  -  sp 

TTT 

5  7 
5 


Figure  3 — Group  6:  Characteristics  and  diagram. 


285 


PLAN  1  2    AXE  1  HORIZONTAL 


AXE  2  VERTICAL 


• 

* 

• 

GBC  I 

• 

BMC  * 

GCE 

• 

• 

• 

• 

I 

• 

• 

• 

« 

- 

\ 

*  RGC 

;  ST3 

• 

• 

ALT 

* 

• 

•  ITC 

* 

• 

* 

•  RGT. 

*       PBC  * 

VEG 


"STV 


,,(|ES, 
COU 


TTE 

GEP 


RGE 


ALT 

altitude 

VEG 

vegetation 

COU 

canopy  of  strata  1 ,  2  and  3  in  1 .10"^ 

GES 

sample  basal  area 

RGT 

total  regeneration  (spruce  and  cembran  pine) 

ST1 

Importance  of  stratum  1  (18-24  m)  in  1.10'^ 

ST3 

importance  of  stratum  3  (06-12  m)  in  1 .10"' 

PBC 

ratio  of  small-size  trees  calculated  (10  cm  <  (|)  < 

20  cm)  in  1.10"' 

BMC 

ratio  of  medium-size  trees  calculated  (25  cm  <  ( 

)  <  35  cm)  in  1.10" 

GBC 

ratio  of  large-size  trees  calculated    >  40  cm)  in  1.10"' 

RGE 

spruce  regeneration  in  1.10"' 

TTE 

number  of  spruce  trees  over  5  ares 

GEP 

spruce  basal  area 

RGC 

cembran  pine  regeneration  in  1.10"' 

TTC 

number  of  cembran  pine  trees  over  5  ares 

GCE 

cembran  pine  basal  area 

Figure  4 — Variables  correlation  circle  (Factorial  Discriminant  Analysis). 


In  figure  4,  the  first  principal  axis  represents  the  grow- 
ing stock: 

•  In  the  positive  part  of  the  axis,  it  is  closely  linked  to 
the  total  number  of  spruce,  to  the  spruce  basal  area,  to  the 
total  canopy,  and  to  the  total  sample  basal  area. 

•  In  the  negative  part  of  the  axis,  it  is  closely  linked  to 
the  cembran  pine  regeneration  and  to  stratum  3  (6  to  12  m) 
(Leclerc  1992). 

The  second  principal  axis  represents  the  individual 
height  of  trees: 

•  In  the  positive  part  of  the  axis,  it  is  strongly  correlated 
with  the  ratio  of  large-size  trees  and  of  medium-size  trees 
as  well  as  with  the  cembran  pine  basal  area. 


•  In  the  negative  part  of  the  axis,  it  is  strongly  correlated 
with  the  small-size  trees  ratio  (Leclerc  1992). 

In  figure  5,  we  can  observe  the  position  of  individuals  as 
well  as  that  of  the  six  groups.  Ninety-six  percent  of  the  in- 
dividuals have  been  well  classified  in  the  factorial  discrimi- 
nant analysis. 

In  the  forest  management  plan  of  Tueda,  Pignol  (1992) 
showed  that  the  mixed  spruce  and  cembran  pine  stands 
had  a  canopy  ranging  from  5  to  40  percent  in  only  12  per- 
cent of  cases — their  canopy  was  over  40  percent  otherwise. 
On  the  other  hand,  in  the  pure  cembran  pine  stands  that 
we  have  not  fully  studied  yet,  the  opposite  can  be  observed: 
over  40  percent  of  the  canopy  can  be  foimd  in  25  percent  of 
the  area. 


286 


So,  in  the  cembran  pine  forest  of  Tueda,  we  can  find 
analogies  with  the  other  French  cembran  pine  stands. 
Yet  it  presents  notable  characteristics,  among  which  is 
its  youth. 

New  And  Future  Studies 

In  our  study  of  mixed  spruce  and  cembran  pine  stands, 
we  have  not  yet  been  able  to  analyze  all  the  data  concern- 
ing the  growth,  age,  and  crown  shape.  We  intend  to  do  so 
in  the  near  future.  Moreover,  the  study  of  pure  cembran 
pine  stands  will  be  carried  out  according  to  a  protocol  simi- 
lar to  the  one  used  for  the  study  of  mixed  stands.  Perma- 
nent inventory  plots  could  be  set  up  in  both  types  of  stands. 

The  closing  of  openings  by  spruce  and  cembran  pine  seed- 
lings will  require  a  grid  inventory  so  as  to  better  determine 
the  dynamics  of  both  the  closing  of  present  openings  and 
the  progression  of  the  timberline.  We  are  considering  com- 
pleting this  work  with  a  dendrochronological  study,  which 
coiild  provide  us  with  some  information  decisive  to  the  un- 
derstanding of  the  cembran  pine  stand  d3Tiamics.  Such  a 
study  could  also  help  us  to  better  understand  the  long-term 
evolution  of  forest  ecosystems  under  the  influence  of  envi- 
ronmental changes. 

A  paleoecological  study  is  under  way.  It  is  being  carried 
out  by  the  Vanoise  Natural  Park.  The  University  of 
Chambery  has  begim  a  study  on  the  cembran  pine  cone 
production  and  its  seed  propagation.  In  addition,  they 
are  considering  carrying  out  a  historical  study  of  the  use 
of  space,  using  the  different  cadastres  available.  They  also 
intend  to  study  the  carbon  stocking  modification  in  the 
different  soil  horizons  in  the  case  of  an  upward  progression 
of  the  timberline. 

Finally,  we  intend  to  analyze  the  possible  correlation  be- 
tween the  biodiversity  indicators  and  the  different  struc- 
tures we  have  observed. 


All  these  studies  should  enable  us  to  have  a  better 
knowledge  of  the  Tueda  cembran  forest  and  of  its  evolu- 
tion mechanisms,  and  they  should  tell  us  if  maintaining 
diversity  requires  some  form  of  management,  and,  if  so, 
which  one. 

CONCLUSIONS 

We  have  noted  the  Tueda  cembran  pine  stand  presents 
very  distinct  types  of  structure.  In  some,  cembran  pine 
is  predominant,  in  others  spruce  is.  Some  have  a  sparse 
canopy,  others  a  very  dense  one.  Regeneration  can  be  ei- 
ther nonexistent  or  abundant,  with  a  predominance  of  ei- 
ther cembran  pine  or  spruce.  Six  stand  groups  have  been 
determined.  What  matters  to  us  now  is  to  have  a  better 
knowledge  of  their  evolution  and  dynamism  so  as  to  man- 
age them  as  well  as  possible.  We  are  considering  carry- 
ing out  some  additional  studies. 

It  is  likely  that  cembran  pine  requires  a  more  energetic 
form  of  silviculture  than  the  one  we  consider  today  in  the 
lower  part  of  its  belt  (lower  subalpine  and  upper  montane 
belt),  where  it  is  in  competition  with  spruce.  The  aim  of 
this  silviculture  is  of  course  to  favor  cembran  pine  develop- 
ment, to  maintain  a  structure  diversity,  and  to  give  it  a 
place  of  choice,  protecting  it,  among  other  things,  against 
possible  domination  by  spruce. 

In  the  middle  subalpine  belt  a  minimal  silviculture,  or 
even  no  silviculture  at  all,  would  be  better  adapted  to  these 
altitudes  where  the  stands  are  very  stable,  irregular,  and 
of  very  slow  growth.  Other  sectors  like  the  station  ofLin- 
nea  borealis  would  require  specific  protection. 

Practices  determined  according  to  the  dynamism  spe- 
cific to  each  ecological  system  will  enable  us  to  enhance 
the  alpine  landscapes  and  to  restore  their  authenticity, 
while  maintaining,  if  not  increasing,  their  biodiversity. 


PLAN    I  2 


AXE  1  HORIZONTAL 


AXE    2  VERTICAL 


Figure  5— Localization  of  individuals  and  groups  (Factorial  Discriminant  Analysis). 


287 


REFERENCES 

Bartoli,  Ch.  1966.  Etudes  ecologiques  sur  les  associations 
forestieres  de  la  Haute  Maurienne  -  Annales  des  sciences 
forestieres  (Tome  XXIII.  Fascicule  3.  1966.  321  p.). 

Combes,  F.  1986.  Le  Pin  cembro  dans  les  Alpes  de  Haute 
Provence  iPinus  cembra  L.).  Revue  forestidre  fran9aise. 
N°2:  135-139. 

Contini,  L.;  Lavarelo,  Y.  1982.  Le  Pin  cembro  iPinus  cembra 
L.),  Repartition,  Ecologie,  Sylviculture  et  Production  - 
Institut  national  de  la  Recherche  agronomique.  197  p. 

Crocq,  C.  1990.  Le  Casse-Noix  mouchete  {Nucifraga  caryo- 
catactes).  326  p. 

Del  Favero,  R.;  De  Mas,  G.;  Lasen,  C.  -  Paiero,  P.  1985.  II 
Pino  cembro  nel  veneto.  91  p. 


Fourchy,  P.  1968.  Notes  sur  le  Pin  cembro  (Pinus  cembra 

L.)  djuis  les  Alpes  fran^aises.  Revue  foresti^re  franpaise. 

N°2  F6vrier:  77-94. 
Inventaire  Forestier  National  Frsinpais.  1986.  Resultats 

du  deuxifeme  inventeiire  forestier.  Depsirtement  de  la 

Savoie.  TOME  1.  142  p. 
Inventaire  Forestier  National  Suisse.  Resultats  du  premier 

inventaire  1982-1986.  375  p.  et  cartes  thematiques. 
Leclerc,  D.  1992.  Traitements  statistiques  des  donn^es  de 

TUEDA.  Office  National  des  Forets.  Section  Technique 

Interr^gionale  des  Alpes  du  Nord.  43  p. 
Ozenda,  P.  1985.  La  vegetation  de  la  chaine  alpine  dans 

I'espace  montagnard  europ6en.  344  p. 
Pignol,  V.  1992.  Procfes- verbal  d'sunenagement  de  la  foret 

communale  de  LES  ALLUES.  Massif  de  Tueda.  1992- 

2002.  Office  National  des  Forets.  63  p. 


288 


Forest  Management 


International  Workshop 
St.  Moritz  1 992 


SILVICULTURAL  TREATMENT  AND 
AVALANCHE  PROTECTION  OF  SWISS 
STONE  PINE  FORESTS 

Werner  Frey 


Abstract — Subalpine  European  larch  (Larix  dec/dua )-Swis8 
stone  pine  (Pinus  cembra)  forests  show  a  natural  succession  of 
larch  and  stone  pine  that  is  strongly  influenced  by  natural  disas- 
ters. Silvicultural  treatment  has  to  consider  natural  tendencies, 
particularly  retaining  a  permanent  amount  of  larch.  On  the 
other  hand,  stone  pine-dominated  forests  are  more  efficient  for 
avalanche  protection  than  larch-dominated  forests.  The  applica- 
tion of  new  results  of  avalanche  research  on  larch-stone  pine  for- 
ests including  their  silvicultural  treatment  is  discussed. 


This  paper  gives  a  short  overview  on  the  natural  devel- 
opment and  regeneration  of  the  subalpine  European 
larch-Swiss  stone  pine  forests  in  the  central  European 
Alps.  In  these  forests,  the  mutual  amount  of  larch  (Larix 
decidua  Mill.)  and  of  stone  pine  (Pinus  cembra  L.)  is 
strongly  influenced  by  natural  disasters  such  as  ava- 
lanches, snow  load,  and  storms.  The  ability  to  protect,  es- 
pecially against  avalanches,  is  different  in  different  types 
of  the  larch-stone  pine  forest.  This  necessitates  different 
silvicultural  treatments. 

NATURAL  DEVELOPMENT  AND 
REGENERATION 

We  use  the  term  "European  larch-Swiss  stone  pine 
forest"  (Larici-Pinetum  cembrae;  Ellenberg  and  Klotzli 
1972)  for  the  natural  commiinity  of  these  types  of  forests 
throughout  the  central  European  Alps.  Older  names  refer 
mostly  to  Rhododendron-rich  varieties. 

The  natural  development  of  the  European  larch-Swiss 
stone  pine  forests  is  described  by  Auer  (1947),  Campell 
(1955),  Ellenberg  (1978),  Ellenberg  and  KlotzH  (1972), 
Mayer  and  Ott  (1991),  and  other  authors.  The  larch-stone 
pine  forests  develop  only  in  the  more  continental  climate 
in  the  subalpine  zone  of  the  Alps.  The  different  stages 
can  be  explained  by  the  fact  that  larch  is  a  pioneer  species 
and  stone  pine  is  a  climax  species. 

One  main  reason  for  the  different  behavior  of  larch  and 
stone  pine  is  the  difference  in  size  of  seeds:  larch  seeds 
are  windborne  and  have  a  weight  of  about  0.005  g  each; 
stone  pine  seeds  are  heavy  (0.25  g  each)  and  dependent  on 
animals  for  transport.  Germination  in  mineral  soil  is  pos- 
sible for  both  species,  but  there  will  be  far  more  larch 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Werner  Frey  is  Forestry  Engineer  and  Scientist  at  the  Swiss  Federal 
Institute  for  Snow  and  Avalanche  Research,  CH-7260  Davos 
Weissfluhjoch,  Switzerland. 


seeds  available  in  such  places,  and  larch  dominates  the 
initial  stages.  The  period  of  larch  domination  lasts  for 
about  one  century.  With  the  gradual  building  up  of  a  lit- 
ter layer  on  the  forest  floor,  regeneration  becomes  difficult 
for  larch  seedlings  that  cannot  reach  the  mineral  soil  im- 
der  the  duff  layers.  In  such  places,  successful  regenera- 
tion is  possible  only  for  seedlings  of  stone  pine.  Stone 
pine  needles  thicken  the  duff  layer,  and  finally  the  larch- 
stone  pine  forest  t\irns  into  a  pvire  stone  pine  forest.  A 
new  initiation  of  this  process  is  often  produced  by  natural 
disasters  such  as  avalanches,  heavy  snow  load,  and 
storms  that  uproot  stone  pines  thus  exposing  mineral  soil. 

In  the  subalpine  zone,  the  growth  of  stone  pine  is  very 
slow.  Its  dominance  in  the  continental  Alps  is  based 
mainly  on  four  factors:  winter  frost  hardiness,  resistance 
to  winter  desiccation  (Frey  1983),  ability  to  use  the  short 
vegetation  period,  and  a  quite  high  shade  tolerance.  Com- 
pared to  stone  pine,  larch  finds  less  optimal  growth  condi- 
tions in  the  subalpine  zone  of  the  central  Alps:  its  growth 
is  well  adapted  to  the  short  vegetation  period  but  is  com- 
bined with  less  winter  frost  hardiness,  less  resistance  to 
winter  desiccation,  and  a  poor  shade  tolerance. 

If  the  natural  succession  is  not  disturbed  by  humans  or 
by  catastrophic  events,  these  factors  usually  lead  to  the 
development  of  pure  stone  pine  climax  stands. 

AVALANCHE  PROTECTION 

In  mountainous  regions,  forests  are  most  important 
for  avalanche  control  (to  prevent  formation  of  avalanches 
endangering  settlements  or  traffic  routes  at  the  foot  of 
mountain  slopes)  (Frey  and  Salm  1990).  In  the  Upper 
Engadine,  as  in  many  other  European  regions,  cultivation 
of  agricultural  land  in  the  past  resulted  in  the  clearing  of 
sunny  slopes.  Development  of  the  villages  and  their  in- 
creasing dependence  on  roads  induced  a  need  for  ava- 
lanche protection.  Since  forests  are  known  to  be  a  good 
and  sustainable  protection  measure,  many  areas  of  affor- 
estation have  been  established  since  the  19th  century. 
Figures  1  and  2  show  some  aspects  for  the  Upper 
Engadine,  Switzerland.  Data  can  be  found  in  Auer 
(1947),  Bundesamt  fiir  Forstwesen  (1982),  and  Schlatter 
(1935). 

How  does  a  forest  protect  areas  fi'om  avalanches?  Some 
recent  results  of  snow  and  avalanche  research  help  an- 
swer this  question  (Frey  and  Salm  1990;  Gubler  and 
Rychetnik  1991;  Meyer-Grass  and  Schneebeli  1992).  Most 
of  these  results  originate  from  spruce  and  larch  stands 
near  timberline  in  the  region  of  Davos.  The  effect  of  stone 
pine  on  the  formation  of  snow  cover  is  quite  similar  to  the 
effect  of  spruce  (Schiechtl  1973). 


290 


1880    1890     1900    1910     1920     1930     1940    1950     1960  1970 

Decade 


Figure  1 — Project  area  and  time  of  establish- 
ment of  important  afforestation  areas  to  pro- 
tect villages  and  main  roads:  Upper  Engadine 
(see  fig.  2). 


Snow  interception  by  trees  is  a  very  important  factor 
in  preventing  avalanches.  It  has  two  main  effects  on  the 
formation  of  extreme  avalanches:  first,  it  modifies  the  dis- 
tribution and  rate  of  accumvdation  of  new  snow  during 
storms;  second,  between  storms  it  alters  the  old  snow  lay- 
ers including  the  snow  surface. 

Radiation  also  influences  strongly  the  quality  of  the 
snowpack  in  the  forest.  Incoming  shortwave  radiation 
heats  needles,  branches,  and  stems.  Temperatures  rise 


above  the  melting  point,  and  intercepted  snow  falls  into 
the  snowpack.  This  produces  a  strong  pattern  of  distur- 
bance and  prevents  the  formation  of  weak  layers  in  the 
snowpack  that  can  result  in  dangerous  slab  avalanches. 

Deciduous  larch  shows  significant  differences  in  snow 
interception  and  influence  on  radiation  compared  to  ever- 
green spruce  and  stone  pine.  In  relation  to  an  open  field, 
the  incoming  shortwave  radiation  and  longwave  radiation 
loss  amounts  to  only  15-30  percent  for  larch,  but  up  to 
90  percent  for  stone  pine.  Adding  differences  in  snow  in- 
terception, mean  snow  depth  reduction  is  10-30  percent  in 
open  larch  stands,  and  up  to  90  percent  in  dense  stone 
pine  stands.  Mostly  during  and  after  heavy  snow  storms, 
stone  pine  prevents  the  formation  of  weak  layers  in  the 
snowpack  more  than  larch  does. 

Compared  to  equivalent  conditions  in  an  open  field,  dry 
slab  avalanches  in  openings  of  subalpine  forests  only  start 
at  considerably  steeper  (5°)  inclinations.  The  thickness  of 
svirface  slab  in  forest  openings  is  significantly  increased 
(Meyer-Grass  and  Schneebeli  1992).  Both  factors  point 
out  the  protection  ability  of  subalpine  forests. 

The  structure  of  subalpine  forests  is  important  to  ava- 
lanche protection.  Gubler  and  Rychetnik  (1991)  found 
that  extreme  dry-slab  avalanches  may  start  in  openings 
with  downslope  lengths  of  as  little  as  30  m  and  widths  of 
15  m.  Smaller  avalanches  may  start  in  openings  with 
widths  of  only  10  m  (Meyer-Grass  and  Schneebeli  1992). 
These  facts  necessitate  small-scale  management 
techniques. 


Figure  2 — Project  areas  of  afforestation  protecting  villages  and  main  roads  in  the 
Upper  Engadine  established  from  about  1 860  to  1980.  Note  that  these  areas  are 
mostly  southern  exposed. 


291 


The  differences  between  larch-dominated  forests  and 
spruce/stone  pine-dominated  forests  in  avalanche  protec- 
tion can  be  summarized:  (1)  dense  stone  pine  stands,  par- 
ticularly if  multilayered,  effectively  prevent  the  formation 
of  extreme  slab  avalanches;  (2)  open  larch  stands  near 
timberline  do  not  significantly  hinder  the  formation  of  ex- 
treme avalanches.  In  addition  to  these  findings  of  Gubler 
and  Rychetnik  (1991),  Meyer-Grass  and  Schneebeh  (1992) 
observed  that  sufficient  dense  clustered  larch  forests 
give  effective  protection  from  the  formation  of  smaller 
avalanches. 

For  avalanche  protection,  the  silvicultural  goal  must  be 
a  mosaic  of  clustered  stands,  sufficiently  dense,  multilay- 
ered, with  mixed  species,  and  uneven-aged.  Clusters  of 
different  age  and,  therefore,  different  protective  capability 
should  alternate  in  space  (Gubler  and  Rychetnik  1991). 

PROTECTION  FROM  EROSION 

It  is  difficult  to  judge  generally  the  erosion  protection 
value  between  different  tree  species  and  forest  tj^jes. 
Important  factors  such  as  geology  and  precipitation 
change  in  space,  and  soil  and  topography  may  change  on 
an  even  smaller  scale.  However,  for  protection  from  ero- 
sion Mayer  and  Ott  (1991)  characterize  the  species  of 
larch-stone  pine  forests  as  follows:  stone  pine  is  able  to  re- 
tain much  more  precipitation  than  larch  and  stemflow  is 
reduced  by  the  coarse  bark  of  both  larch  and  stone  pine. 
In  rocky  soils  stone  pine  is  deeper  rooting  than  larch.  The 
silvicultural  goal,  therefore,  is  quite  similar  to  the  goal  in 
avalanche-protection  forests:  a  natiiral  mixture  of  larch 
and  stone  pine,  sufficiently  dense,  multilayered,  and 
uneven-aged. 

SILVICULTURAL  TREATMENTS 

In  larch-stone  pine  forests  the  full  succession  from  pio- 
neer stands  rich  in  larch  to  mature  stone  pine  stands 
takes  more  than  400  years  (Mayer  and  Ott  1991).  Suffi- 
cient avalanche  and  erosion  protection  can  be  guaranteed 
only  by  a  small-scale  mosaic  of  multilayered  and  clustered 
forest  types.  Both  natural  development  and  the  objective 
of  protection  eliminate  all  large-scale  silvicultural  treat- 
ment methods. 

Different  silvicultural  treatments  are  used  for  different 
types  of  forest  utilization: 

1.  Grazing  woodland  or  forested  recreation  areas  on 
gentle  slopes  with  little  need  for  protection. 

2.  Forests  on  gentle  or  steeper  slopes  used  for  wood 
production  but  also  for  protection. 

3.  Avalanche-  and  erosion-protection  forests  mostly  on 
steep  slopes. 

Small-Scale  Methods 

Before  talking  about  silvicultural  methods  used  in  the 
different  utilization  types,  we  have  to  introduce  some 
small-scale  silvicultural  treatment  methods  that  are  gen- 
erally used  in  the  European  Alps.  Details  can  be  found  in 
the  publications  of  Auer  (1947),  Aulitzky  and  Turner 
(1982),  Bischoff  (1987),  Campell  (1955),  Mayer  and  Ott 


(1991),  Pitterle  (1988),  Piussi  and  Schneider  (1985),  and 
Trepp  (1981). 

Selection  Method  (single-tree  selection  method;  Ger- 
man term  "Einzelplenterimg") — The  individual  harvesting 
area  of  this  method  is  very  small.  Harvesting  is  realized 
by  felling  single  selected  trees  from  different  stories,  re- 
taining and  creating  favorable  conditions  for  regeneration 
at  the  same  time.  Forests  managed  by  this  method  are 
multilayered  and  clustered  and  show  a  mixture  of  all  tree 
ages  over  very  small  areas.  This  method  as  used  in  the 
past  showed  a  strong  tendency  to  ehminate  larch  in  the 
larch-stone  pine  forest. 

Cluster  Selection  Method  (small-scale,  group- 
selection  method;  German  terms  "Gebirgsplenterung"  or 
"Gruppenplenterung") — The  individual  harvesting  area 
is  quite  small.  In  subalpine  forests,  trees  usually  grow  in 
an  arrangement  of  small  clusters  in  which  trees  mutually 
protect  each  other  (Schonenberger  and  others  1990). 
These  clusters  are  now  considered  as  individual  trees;  the 
group-selection  method  harvesting  whole  tree  clusters  is 
aiming  at  the  same  goal  as  the  single-tree  selection 
method. 

Group  Selection  Method  (German  term:  "Femel- 
schlag") — Harvesting  areas  are  bigger  and  may  vary  from 
one-third  to  several  hectares.  The  local  distribution  of 
successional  stages  and  the  variability  of  sites  define  the 
harvesting  area.  Regeneration  nuclei  and  logging  aspects 
have  to  be  taken  into  consideration. 

Use  of  Methods 

These  silvicultural  methods  in  the  larch-stone  pine  for- 
est are  applied  in  three  types  of  areas: 

1.  Grazing  woodland  or  forested  recreation  areas  on 
gentle  slopes.  The  aim  is  an  open  structure  with  a  good 
portion  of  larch.  These  forests  are  open  to  warmth  and 
light.  Selection  harvesting  of  mature  trees  combined  with 
measures  promoting  natural  regeneration  or  planting  is 
best  suited  to  the  management  objectives.  Young  trees 
have  to  be  protected  against  deer  and  cattle  if  necessary. 

2.  Forests  on  gentle  or  steeper  slopes  grown  primarily 
for  wood  production.  The  natural  succession  is  leading  to 
quite  pure  stone  pine  forests  and  can  be  accelerated  using 
the  single-tree  selection  method.  A  higher  amount  of 
larch  can  be  obtained  using  the  group-selection  method  or 
the  cluster-selection  method,  if  necessary  sustained  by 
soil  stripping  (scarification)  to  prepare  a  good  germination 
bed  for  larch. 

3.  Avalanche  and  erosion  protection  forests,  mostly  on 
steep  slopes.  The  stability  of  these  forests  in  time  and 
space  is  dependent  on  a  sufficient  proportion  of  larch  to 
prevent  the  disintegration  of  overmature  stands  composed 
purely  of  stone  pine.  On  the  other  hand,  stone  pine  is 
more  effective  in  avalanche  and  erosion  protection.  A  per- 
manent protection  effect  needs  siifficient,  well-dispersed 
regeneration.  In  potential  avalanche  starting  zones, 
openings  should  not  exceed  downslope  lengths  of  30  m 
and  widths  of  15  m.  The  cluster-selection  method  best 
fits  these  management  objectives. 


292 


SUMMARY  AND  CONCLUSIONS 

Natxiral  succesion  in  larch-stone  pine  forests  starts  with 
a  high  amoiint  of  larch  germinating  and  growing  well  on 
mineral  soil.  Mature  larch  stands  produce  a  relatively 
shaded  forest  floor  and  a  building  up  of  a  duff  layer  cov- 
ered by  a  needle  layer.  These  conditions  favor  germina- 
tion and  growth  of  stone  pine.  Mature  stands  consist  pre- 
dominantly of  stone  pine.  Natvu^al  succession  is  intitiated 
by  disasters  such  as  avalanches,  heavy  snow  load,  or 
storms.  The  succession  period  will  last  for  more  than 
400  years. 

The  silviciiltural  goal  for  good  avalanche  and  erosion 
protection  must  be  a  natural  mosaic  of  larch  and  stone 
pine  stands  in  a  clustered  structure,  svifficiently  dense, 
multilayered,  and  uneven-aged. 

Experience  has  shown  that  the  best  silvicultural  meth- 
ods for  wood  production  are  the  group-selection  or  the 
cluster-selection  methods  that  allow  the  favoring  of  either 
larch  or  stone  pine  according  to  management  needs.  The 
most  suitable  method  for  protection  purposes  is  the 
cluster-selection  method. 

ACKNOWLEDGMENTS 

Thanks  to  Ian  McCracken,  Walter  Schonenberger, 
Martin  Meyer-Grass,  Walter  Ammann,  and  Reinhard 
Lassig,  who  reviewed  the  manuscript  critically. 

REFERENCES 

Auer,  C.  1947.  Untersuchungen  iiber  die  natiirHche  Ver- 
jiingung  der  Larche  im  Arven-Larchenwald  des  Ober- 
engadins.  Eidg.  Anst.  forstl.  Versuchswes.,  Mitt.  25(1): 
7-140. 

Aulitzky,  H.;  Turner,  H.  1982.  Bioklimatische  Grundlagen 
einer  standortsgemassen  Bewirtschaftung  des  sub- 
alpinen  Larchen-Zirbenwaldes.  Eidg.  Anst.  forstl. 
Versuchswes.,  Mitt.  58(4):  325-580. 

Bischoff,  N.  1987.  Pflege  des  Gebirgswaldes.  Leitfaden  fiir 
die  Begnindung  und  forstliche  Nutzung  von  Gebirgs- 
waldern.  Bern,  Bimdesamt  fiir  Forstwesen  imd  Land- 
schaftsschutz.  379  p. 

Bimdesamt  fiir  Forstwesen.  1982.  Lawinenverbau-  und 
Aufforstungskataster.  Bern. 

Campell,  E.  1955.  Der  Larchen-Arvenwald  (Alpenrosen- 
Heidelbeergesellschaft,  Rhodoreto-Vaccinietima).  In: 
Ertragreiche  Nadelwaldgesellschaften  im  Gebiete  der 
schweizerischen  Alpen...  von  E.  Campell,  R.  Kuoch,  F. 
Richard,  W.  Trepp.  Biindnerwald,  Beiheft  5:  14-26. 


Ellenberg,  H.  1978.  Vegetation  Mitteleuropas  mit  den 

Alpen  in  okologischer  Sicht.  Verlag  Eugen  Ulmer, 

Stuttgart.  982  p. 
Ellenberg,  H.;  KlbtzH,  F.  1972.  Waldgesellschaften  und 

Waldstandorte  in  der  Schweiz.  Eidg.  Anst.  forstl. 

Versuchswes.,  Mitt.  48(4):  589-929. 
Frey,  W.  1983.  The  influence  of  snow  on  growth  and 

survival  of  planted  trees.  Arctic  and  Alpine  Research. 

15(2):  241-251. 
Frey,  W.;  Salm,  B.  1990.  Snow  properties  and  movements 

in  forests  of  different  climatic  regions.  In:  Proc.  XIX 

lUFRO  World  Congress,  Montreal,  Canada,  Div.  1,  Vol. 

1:  328-339. 

Gubler,  H.;  Rychetnik,  J.  1991.  Effects  of  forests  near  tim- 
berHne  in  avalanche  formation.  In:  Bergmann,  H.; 
Lang,  H.;  Frey,  W.;  Issler,  D.;  Salm,  B.,  eds.  Snow,  hy- 
drology and  forests  in  high  alpine  areas.  lAHS  Publ. 
202:  19-38. 

Mayer,  H.;  Ott,  E.  1991.  Gebirgswaldbau  -  Schutz- 
waldpflege:  Ein  waldbaulicher  Beitrag  zur  Land- 
schaftsokologie  imd  zimi  Umweltschutz.  2.  Auflage. 
Gustav  Fischer  Verlag,  Stuttgart.  587  p. 

Meyer-Grass,  M.;  Schneebeli,  M.  1992.  Die  Abhangigkeit 
der  Waldlawinen  von  Standorts-,  Bestandes-  imd 
Schneeverhaltnissen.  In:  Schutz  des  Lebensraumes  vor 
Hochwasser,  Muren  imd  Lawinen.  Interpravent,  Bern. 
2:  443-445. 

Pitterle,  A.  1988.  Waldbauliche  Analyse  und  Behand- 
lungsmassnahmen  von  anthropogen  beeinflussten 
subalpinen  Fichten-  sowie  Larchen-Zirbenwaldern  im 
Villgratental/Osttirol.  Diss.  Univ.  Bodenkultur,  Wien. 
225  p. 

Piussi,  P.;  Schneider,  A.  1985.  Die  obere  Wald-  und  Baum- 
grenze  im  Pfitschtal  (Siidtirol).  [Upper  forest  limits  and 
treelines  in  the  Pfitschtal,  South  Tyrol].  Cbl.  ges. 
Forstw.  102(4):  234-246. 

Schiechtl,  H.  M.  1973.  Wiederbewaldung  von  Extrem- 
standorten  -  Grundlagen  und  Voraussetzungen  in  den 
Hochlagen  und  auf  Rohbbden.  Allgem.  Forstztg.,  Wien. 
84(10):  243-245. 

Schlatter,  A.  J.  1935.  Die  Aufforstungen  und 
Verbauungen  des  Oberengadins  in  den  Jahren  1875- 
1934.  Schweiz.  Z.  Forstwes.  86. 

Schonenberger,  W.;  Frey,  W.;  Leuenberger,  F.  1990. 
Okologie  und  Technik  der  Aufforstung  im  Gebirge  - 
Anregungen  fiir  die  Praxis.  Eidg.  Anst.  forstl. 
Versuchswes.,  Ber.  325.  58  p. 

Trepp,  W.  1981.  Das  Besondere  des  Plentems  im 
Gebirgswald.  Schweiz.  Z.  Forstw.  132(10):  823-846. 


293 


IMPORTANCE  AND  SILVICULTURAL 
TREATMENT  OF  STONE  PINE  IN  THE 
UPPER  ENGADINE  (ORISONS) 

Riet  Gordon 


Abstract — The  European  larch  (Larix  decidua)-SvfisB  stone  pine 
(Pinus  cembra)  forests  of  Celerina  have  been  affected  by  humans 
for  centuries.  Today's  demands  on  the  forest  have  an  important 
influence  on  the  silvicultural  treatment  of  these  forests.  The 
stone  pine,  as  the  naturally  best-adapted  tree  species  in  Celerina, 
suffers  from  the  consequences  of  public  demands.  The  selective 
cutting  of  stone  pine  often  appears  necessary  to  promote  the 
Scots  pine  and  larch  as  the  most  important  tree  species  for  con- 
servation and  recreation  functions.  Nevertheless,  in  the  future 
stone  pine  will  continue  to  be  the  most  important  tree  species  in 
the  forests  of  the  Upper  Engadine. 


Most  of  the  alpine  forests  have  been  affected  by  humans 
for  centuries.  The  European  larch  {Larix  decidua)-Swiss 
stone  pine  (Pinus  cembra)  forest  in  the  Upper  Engadine  is 
no  exception.  Human  influence  is  apparent  in  the  forest 
structure  and  species  composition  (Ganzoni  1911).  In  re- 
cent times,  the  demands  on  the  forest  have  changed.  Thus 
forest  management  and  treatment  have  also  changed. 

This  paper  discusses  the  management  and  treatment 
of  stone  pine  forests  in  an  area  often  used  by  the  public. 
Three  examples  from  the  forest  enterprise  in  Celerina  are 
presented. 

HISTORY  OF  CELERINA'S  FORESTS 

Until  the  end  of  last  century,  the  forests  in  Celerina 
were  used  for  wood  production  for  local  purposes  and  for 
grazing  indigenous  cattle  and  enormous  herds  of  sheep 
from  Italy.  These  sheep  used  to  spend  the  summer  in  the 
Upper  Engadine  (Schlatter  1935).  In  addition,  forest  lit- 
ter was  used  as  bedding  in  the  cattle  stables,  and  in  times 
of  famine  stone  pine  seeds  even  served  as  an  additional 
source  of  food  for  the  local  people. 

The  village  constitution  includes  forest  laws  dating  back 
to  1699  (Ganzoni  1982).  The  evolution  of  these  laws  dem- 
onstrates the  importance  of  the  forests  to  the  village  com- 
munity. Only  the  village  community  as  a  whole  had  the 
right  to  use  the  forest,  not  the  individual  inhabitant.  To 
enforce  these  laws,  the  village  employed  forest  guards. 
The  wood  was  normally  used  by  the  village  community, 
and  only  in  special  cases  was  it  sold.  The  inhabitants  of 
the  village  were  entitled  to  a  limited  amount  of  building 
wood  and  firewood,  but  only  for  their  own  personal  use. 
They  were  also  allowed  to  collect  litter,  peat,  or  stone  pine 
cones. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Riet  Gordon  is  Forestry  Engineer  and  Scientist,  Swiss  Federal  Institute 
of  Technology,  Department  of  Forestry,  CH-8092  Zurich,  Switzerland. 


During  the  17th  and  18th  centuries,  the  exploitation 
of  some  forests  in  Celerina  was  prohibited.  The  preserved 
forests  were  not  "protection  forests"  as  in  many  places  in 
the  Alps;  they  were  probably  forests  that  had  been  over 
used.  The  ban  on  exploitation  can  be  considered  as  the 
beginning  of  forest  management. 

With  the  exception  of  the  preserved  forests,  all  forests 
had  been  intensively  pastured  until  the  end  of  the  19th 
century.  The  first  regulations  concerning  pasturing  date 
fi'om  1860.  From  then  on,  grazing  in  the  forests  of  Celerina 
decreased,  except  during  times  of  economic  crises  when 
the  forests  were  intensively  pastured  again. 

At  the  beginning  of  the  present  century,  the  forests  of 
Celerina  became  important  as  a  recreation  area  for  the 
local  people  and  tourists.  Diuing  the  first  decade,  the  for- 
ests were  frequented  mostly  in  summer  by  a  few  tourists. 
Today,  they  are  frequented  by  thousands  of  people  for  hik- 
ing, riding  bicycles,  skiing,  and  horseback  riding  through- 
out the  year.  The  first  forest  management  plan  in  1888 
already  included  some  notes  concerning  the  recreation 
function  of  the  forest.  It  was  proposed  to  establish  more 
and  better  trails  and  to  keep  them  clean  of  branches  and 
fallen  trees.  This  illustrates  the  importance  of  the  forests 
to  tourism  as  early  as  the  beginning  of  the  20th  century. 

In  the  Swiss  lowland  forests  many  plants  and  animals 
became  extinct  (Schweiz  Bund  f.  Naturschutz  1992).  In 
contrast  to  the  lowland  forests,  the  mountain  forests  are 
still  quite  natural.  Therefore  they  are  increasingly  impor- 
tant for  nature  and  landscape  conservation.  In  Celerina, 
conservation  in  the  forest  gained  some  importance  when 
the  marvelous  upland  moors  in  the  "Stazenvald"  were  put 
under  protection  in  1956. 

How  does  this  evolution  influence  the  occurrence  and 
the  treatment  of  the  stone  pine?  Larch  was  the  most  im- 
portant species  in  the  forests  of  Celerina  until  1970,  espe- 
cially on  the  south-facing  slopes.  Larch  produces  very 
good  and  durable  wood  for  construction.  It  was  especially 
used  for  houses  and  barns  but  also  for  artificial  barriers 
to  prevent  floods  and  soil  erosion,  for  water  pipes,  etc. 
The  open  larch  forests  were  suitable  for  grazing  by  cattle 
and  sheep  and  litter  production  for  the  farmers. 

For  a  long  time  stone  pine  was  used  only  for  firewood 
or  as  construction  wood  of  minor  quality.  In  the  dark 
stone  pine  forests  the  pasturage  did  not  have  a  very  high 
production  level  and  litter  production  was  lower  than  in 
larch  forests.  Therefore,  the  management  gave  stone  pine 
a  low  priority  compared  to  larch. 

The  situation  today  has  changed.  The  forest  enterprise 
is  based  financially  on  the  valuable  stone  pine  wood.  Its 
value  is  nearly  four  times  greater  than  that  of  spruce  or 
larch  wood. 


294 


FOREST  MANAGEMENT 

Since  1888,  the  forest  of  Celerina  has  been  cultivated 
on  the  basis  of  a  management  plan.  The  objective  of  this 
management  has  always  been  the  quantitative  preserva- 
tion of  the  forest.  At  the  end  of  the  19th  century,  most  of 
the  mo\intain  forests  were  heavily  damaged  by  grazing, 
wood  over-exploitation,  avalanches,  and  erosion.  The  man- 
agement was  mainly  focused  on  wood  production.  Through 
periodic  inventories,  first  by  the  full  enumeration  method 
and  later  by  sample  plots,  the  state  and  the  development 
of  the  growing  stock,  stem  count  distribution,  species  com- 
position, and  growth  increment  have  been  determined. 
The  result  of  the  planning  process  was  the  prescribed  yield, 
which  was  compulsory  for  10-20  years. 

A  kind  of  silvicultural  planning  already  existed  in  the 
first  management  plan.  It  was  based  mainly  on  natural 
sites.  The  functions  of  the  forest  were  not  yet  considered. 

The  forest  management  was  performed  by  the  State  For- 
est Service  and  not  by  the  forest  owner. 

Management  in  the  Future 

In  the  beginning  of  this  century,  quantitative  forest  con- 
servation was  the  main  goal  of  forest  policy,  a  goal  that  has 
been  achieved  (Bachmann  1991).  Today,  qualitative  forest 
conservation  is  the  main  aim  of  forest  policy.  The  forest 
management  has  to  follow  this  new  objective  to  be  credible 
for  the  Forest  Service,  the  forest  owner,  and  the  general 
public.  What  are  the  needs  for  the  future? 

Elaborating  the  general  aims  of  the  forest  enterprise  is 
not  an  exclusive  affair  of  the  State  Forest  Service,  because 
the  interest  of  the  public  in  the  forest  is  becoming  increas- 
ingly important.  The  forest  owner  and  especially  the  pub- 
lic have  to  be  integrated  in  the  target  process,  to  consider 
their  demands  for  the  forests. 

The  Forest  Service  is  no  longer  the  only  institution  where 
decisions  concerning  the  forests  are  made.  More  and  more 
it  takes  charge  of  consultative  and  coordinative  functions. 
The  decisions  are  made  in  a  democratic  process  open  to  the 
general  public.  This  does  not  mean  that  every  demand  of 
the  public  on  the  forests  has  to  be  accepted  and  realized. 
The  qualitative  preservation  of  the  forest  is  still  the  main 
aim  and,  of  course,  it  has  priority  in  case  of  conflicts  be- 
tween the  demands  and  forest  preservation.  The  results  of 
the  target  process  are  presented  and  illustrated  on  a  forest 
function  map  (fig.  1). 

Information  Needs 

Management  and  silviculture  treatment  have  to  be  har- 
monized with  the  forest  functions.  The  need  for  informa- 
tion is  related  to  the  forest  functions.  For  the  mountain 
forest  the  following  information  is  of  primary  importance: 

1.  Stand  description,  which  consists  of  the  development 
stage,  canopy  closure,  structure,  and  species  composition. 

2.  Stability,  which  shows  the  ability  of  a  forest  stand  to 
remain  constant  or  to  persist  in  the  face  of  disturbance  fac- 
tors such  as  wind,  snow,  avalanches,  and  soil  erosion.  Sta- 
bility is  described  by  the  following  main  features:  regenera- 
tion, structure,  and  density  of  the  stand,  canopy  size,  root 
anchorage,  and  damage  to  the  single  trees  (Ott  1988). 


3.  The  ability  of  a  forest  stand  to  continuously  provide 
all  the  demands  allocated  to  it  by  main  function  suitability. 

The  inventory  methods  have  to  be  adapted  to  this  need 
for  information.  The  present  sample  inventory  is  there- 
fore replaced  or  enhanced  by  a  stand-based  inventory. 

Some  examples  of  the  management  plan  of  Celerina 
illustrate  the  silvicultural  consequences  due  to  forest 
functions. 

PROTECTION  FORESTS 

The  forests  of  Celerina  have  a  relatively  small  direct 
protection  function  compared  to  other  movmtain  forests. 
Nevertheless,  some  forests  with  a  particular  protection 
function  have  been  determined.  For  these  forests  the  fol- 
lowing silvicultural  objectives  have  been  fixed: 

1.  A  small-scale  mosaic  of  all  size  and  age  classes  (clus- 
ter selection  forest). 

2.  At  least  60  percent  stone  pine,  less  than  40  percent 
larch. 

3.  Stable  forest  communities. 

This  aim  can  be  achieved  only  by  regular  silvicultural 
treatment.  Stone  pines  must  be  promoted.  Damage  to 
the  remaining  stand  due  to  logging  or  other  influences 
must  be  avoided.  Therefore,  a  minimal  basic  network 
of  truck  roads  is  needed  in  some  places  to  enable  logging 
with  mobile  cable  cranes.  Other  forest  functions  have  to 
be  subordinated  to  the  protective  function  of  the  forest. 
To  solve  conflicts  in  the  protection  forests  in  Celerina  the 
following  general  measures  are  necessary: 

1.  Preventing  deer  damage  to  trees  by  reducing  the 
number  of  deer. 

2.  Excluding  pasture  from  the  protection  forest. 

3.  Regulating  "wild"  skiing  outside  the  official  skiing 
runs. 

More  information  about  the  avalanche  protection  func- 
tion of  the  forest  and  silvicultural  treatment  of  Eiiropean 
larch-Swiss  stone  pine  forests  was  presented  by  Frey 
(these  proceedings). 

RECREATION  FORESTS 

Recreation  is  of  great  importance  in  the  forests  of 
Celerina.  Recreation  forests  should  have  a  diversified 
structure,  density,  and  species  composition.  Existing  clear- 
ings have  to  be  preserved  as  well  as  special  trees,  particu- 
lar sites,  etc.  The  visual  image  of  the  recreational  forest 
in  Celerina  is  more  important  than  its  wood  production. 

Recreation  function  requires  the  promotion  of  larch  and 
Scots  pine  {Pinus  syluestris)  instead  of  stone  pines.  Larch 
and  Scots  pine  are  pioneer  trees  and  therefore  are  difficult 
to  regenerate  in  closed  forests.  Special  regeneration  tech- 
niques such  as  larger  regeneration  cuts  than  usual  and 
soil  scrapes  have  to  be  applied.  A  further  measure  in  rec- 
reation forests  is  to  cut  old  trees  later  than  proposed  by 
3deld  models. 

Recreation  forests  need  a  special  infrastructure.  In 
summer,  the  forests  of  Celerina  must  have  trails,  horse 
trails,  bicycle  paths,  picnic  and  fire  places,  fitness  courses, 
and  an  information  trail  explaining  the  forests  as  an  eco- 
system. In  winter,  walking  paths,  cross-country  skiing 


295 


Figure  1— Forest  functions  in  the  Celerina  area  of  the  Upper  Engadine  in  Switzerland. 


296 


Table  1 — The  different  functions  and  their  consequences  for  the  silvicultural  treatment 


Function 

Stone  pine 

Larcli 

Structure 

Density 

Yield 

Remarks 

_  _  _  _  _  _Por/^orif  - 

r\yv 

Multifunction 

60 

40 

Clustered 

Normal 

200 

Protection 

>60 

<40 

Clustered 

Closed 

<200 

No  damage,  no  big  openings 

Recreation 

60 

40 

Clustered 

Normal 

>200 

Infrastructure 

Conservation 

>60 

40 

Clustered 

Normal 

>200 

Old  trees,  clearings,  any 

(capercaiilie) 

Open 

accessibility,  silence 

Pasture 

20 

>60 

Open 

>200 

Wood  production 

>60 

<40 

Clustered 

Normal 

200 

Accessibility 

tracks,  and  downhill  ski  runs  have  to  be  maintained.  The 
Forest  Service  is  not  the  only  authority  responsible  for  the 
infrastructure,  but  nevertheless  it  has  an  important  coor- 
dinating function. 

The  recreation  function  involves  some  conflicts  espe- 
cially with  regard  to  nature  conservation  for  species  such 
as  a  large  grouse  in  the  area,  the  capercaiilie  (Terao  uro- 
gallus),  hunters  (disturbance  of  deer),  and  forest  accessi- 
bility. These  conflicts  have  been  smoothed  by  separating 
the  different  demands  on  the  forest  in  time  and  space. 

CONSERVATION  FORESTS 

For  nature  conservation  two  factors  are  important:  pro- 
tection of  moorlands  (bogs)  and  the  habitat  of  the  caper- 
caiilie. The  preservation  of  moorland  demands  that,  due 
to  natural  regeneration  of  yoxmg  stone  pines  and  mountain 
pines  {Pinus  montana),  young  trees  have  to  be  removed 
continuously.  Other  silvicultural  or  technical  measures 
are  neither  necessary  nor  desired.  Bog  areas  should  be 
kept  free  of  paths,  cross-country  tracks,  and  drainage 
ditches.  There  are  only  a  few  areas  left  today  that  are  not 
influenced  by  these  impacts,  because  the  most  important 
cross-country  track  in  the  area  goes  through  the  most 
beautiful  and  largest  bog. 

The  preservation  of  the  "Stazerwald"  as  an  important 
habitat  for  capercaillies  is  considered  the  second  essential 
objective  of  conservation.  Therefore  the  forester  has  to 
favor  Scots  pine,  a  large  nimiber  of  old  trees,  and  sufficient 
open  areas  within  the  forest. 

The  preservation  of  Scots  pine  is  quite  difficult.  Natural 
succession  goes  from  Scots  pine  to  stone  pine.  This 
development  has  to  be  stopped  by  favoring  the  natural 
regeneration  of  Scots  pine  through  exposing  the  mineral 
soil.  Often  this  measure  is  not  sufficient  to  enhance  Scots 
pine  regeneration,  and  additional  planting  is  necessary. 

Besides  these  silvicultural  measures,  other  precautions 
are  necessary  to  preserve  the  capercaiilie  in  the  "Stazer- 
wald." Dviring  mating  and  breeding  time  and  in  winter, 
the  territory  of  the  capercaiilie  should  not  be  disturbed 
by  humans.  For  this  reason,  silvicultural  work  should 
be  delayed  until  after  mid-July.  Before  that  not  even  tour- 
ists should  cross  these  habitats.  Therefore,  some  walking 


paths  and  cross-country  tracks  have  to  be  closed  to  the 
public.  Trespassing  on  the  forests  outside  walking  paths 
is  restrained  until  the  middle  of  July. 

CONCLUSIONS 

The  different  demands  of  the  public  on  the  forests  have 
to  be  taken  into  consideration  during  planning.  Silvicul- 
tural objectives  and  measures,  control,  and  need  for  infor- 
mation (Informationsbedarf)  depend  on  the  forest  functions. 

Considering  the  public  demands  on  the  forest,  the  Swiss 
Forest  Service  nevertheless  uses  multifunctional  forest 
management.   However,  many  specific  demands,  well  de- 
fined in  time  and  space,  can  only  be  achieved  if  the  treat- 
ment is  adapted  to  these  special  objectives.  It  is  not  pos- 
sible to  achieve  these  objectives  only  by  wood  production. 

The  consequences  for  silvicultural  treatment  of  the  for- 
est of  Celerina  are  shown  in  table  1. 

The  stone  pine,  as  the  naturally  best-adapted  tree  spe- 
cies in  Celerina,  is  most  affected  by  public  demands  be- 
cause it  is  mainly  stone  pine  that  has  to  be  "cut  back." 
Nevertheless,  in  the  future  stone  pine  will  continue  to 
be  the  most  important  tree  species  in  the  forests  of  the 
Upper  Engadine. 

REFERENCES 

Bachmann,  P.  1990.  Forsteinrichtung  und  Walderhal- 

tung.  Schweiz.  Z.  Forstwes.  141(6):  415-430. 
Ganzoni,  G.  P.  1982.  Monografia  da  Schlarigna.  Uniun 

dais  Grischs,  Schlarigna. 
Ganzoni,  Z.  1911.  Die  Waldungen  des  Oberengadins. 

Schweiz.  Z.  Forstwes.  62:  40-44,  77-81. 
Ott,  E.;  Schoenbaechler,  D.  1988.  Stabilitaetsbeurteilung 

im  Gebirgswald  als  Voraussetzimg  fuer  die  Schutzwald 

Ueberwachung  und  Pflege  Schweiz.  Z.  Forstwes. 

137(1986):  725-738. 
Schlatter,  A.  J.  1935.  Die  Auforstimgen  und  Verbauungen 

des  Oberengadins  in  den  Jahren  1875-1934.  Schweiz.  Z. 

Forstwes.  86(1935)9:  309-328. 
Schweizer  Bund  f.  Naturschutz.  1992.  Naturwald. 

Sonderheft  Nr.  3,  Basel. 


297 


CULTIVATION  OF  CEMBRAN  PINE 
PLANTS  FOR  HIGH-ELEVATION 
AFFORESTATIONS 

Jorg  Heumader 


Abstract — In  this  paper  a  special  "biological"  method  of  cultivat- 
ing cembran  pine  {Pinus  cembra)  plants  for  high-elevation  affor- 
estations is  described.  This  method  produces  fully  mycorrhized 
transplants  and  pot  plants,  which  have  proved  to  be  well  adapted 
to  the  harsh  environment  of  subalpine  and  timberline  Eireas  in 
the  Inner  Alps  of  Austria. 


The  Alps  are  probably  the  most  intensively  settled  and 
used  high-mountain  range  in  the  world.  Originally,  the 
economic  base  of  the  settlements  in  the  Alps  was  live- 
stock, and  a  special  ranching  and  grazing  system  was  de- 
veloped. Large  areas  in  the  timberline  region  were  defor- 
ested for  livestock  grazing  in  summer  and  for  producing 
hay  for  the  winter  time. 

Natural  hazards  are  "normal"  phenomena  in  high- 
mountain  ranges — with  and  without  anthropogenic  influ- 
ences. On  some  sites,  deforestation  by  humans  had  nega- 
tive effects.  Deforesting  imstable  mountain  sides  or  steep 
slopes  sometimes  led  to  and  sometimes  increased  torrent, 
avalanche,  or  rockfall  disasters. 

One  of  the  tasks  of  the  Austrian  Federal  Service  for 
Torrent  and  Avalanche  Control,  therefore,  is  planning  and 
conducting  high-elevation  afforestations  in  subalpine  ar- 
eas deforested  long  ago  (fig.  1).  High-elevation  afforesta- 
tion in  the  Tyrolean  Alps  got  a  strong  impetus  from  the 
avalanche  catastrophe  of  1951.  Many  disaster-prevention 
projects  with  integrated  control  measures  were  planned, 
and  many  plants  adapted  to  timberline  conditions  were 
needed,  especially  cembran  pine  (Pinus  cembra). 

Cembran  pine  plants  were  hard  to  get  and  not  much  was 
known  about  their  cultivation  at  that  time.  Therefore,  the 
Federal  Service  for  Torrent  and  Avalanche  Control  in 
T)T*ol  not  only  started  a  research  program  on  subalpine 
forests  but  also  founded  in  1953  the  tree  nursery  "Klaus- 
boden"  for  the  production  of  native  subalpine  tree  species. 

CULTIVATION  PRINCIPLES 

The  cultivation  method  for  cembran  pine  plants  was  to 
be  developed  to  a  high  standard  in  this  nursery  in  close 
cooperation  vnth  the  Federal  Forest  Research  Institute 
(Heumader  and  Gobi  1990;  Leys  1970). 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Jorg  Heumader  is  Graduated  Engineer  of  Forestry,  Federal  Service  for 
Torrent  and  Avalanche  Control,  Langgasse  88,  A-6460  Imst,  Tirol,  Austria. 


The  aim  from  the  beginning  was  the  production  of 
plants  of  high  quality  and  special  fitness  for  subalpine 
conditions.  Research  had  shown  that  mycorrhizae  are  es- 
sential for  the  growth  and  surviv£d  of  cembran  pine  plants 
(fig.  2),  because  unmycorrhized  plants  are  not  able  to  get 
enough  nutrients  from  the  thick  raw-hvunus  layers  in  sub- 
alpine areas  (Gobi  1965). 

To  produce  vital  and  mycorrhized  plants  (fig.  3),  a 
special  treatment  of  soil  in  the  nursery  and  a  biological 
cultivation  method  were  developed  with  the  following 
characteristics: 

•  Mycorrhizae  inoculation  of  seed  and  transplant  beds. 

•  Enrichment  of  soil  by  compost  and  peat  to  obtEun  best 
soil  conditions  for  mycorrhizae  growth  with  pH  val- 
ues between  6.0  and  6.5. 

•  Green  manuring  instead  of  chemical  fertilizers, 
which  means  that  every  fourth  year  each  bed  is 
planted  with  a  special  mixture  of  annual  plants  with 
a  high  percentage  of  legumes  for  soil  regeneration 
(Czell  and  Redhch  1966). 

•  No  use  of  herbicides;  weeding  is  done  only  hand. 

MYCORRHIZAE  INOCULATION 

For  the  production  of  mycorrhized  plants,  special  tech- 
niques had  to  be  developed  (Gobi  1979;  Moser  1958,  1959). 


Figure  1 — High-elevation  reforestation  in  the 
Tyrolean  Alps  with  cembran  pine  and  larch 
(Larix  decidua)  on  a  mountain  side  deforested 
for  livestock  grazing  more  than  1 ,000  years  ago. 


298 


Figure  2 — Fine  roots  of  a  cembran  pine  plant  with 
fully  developed  mycorrhizae. 


The  selection  of  smtable  fungi  was  based  on  investiga- 
tions in  cembran  pine  forests,  especially  on  sites  with 
good  natural  regeneration.  The  two  species  Suillus 
placidus  and  Suillus  plorans  were  foiind  to  be  best  of 
all  mycorrhizal  fungi  growing  in  timberline  areas.  On 


Figure  3 — Two  cembran  pine  pot  plants  of  the 
same  age.  The  better  condition  and  growth  of 
the  mycorrhized  plant  on  the  left  is  clearly  seen, 


natural  sites  cembran  pine  mycorrhizae  prefer  the  raw- 
humus  layers  of  podzolic  soils;  corresponding  to  this,  peat 
was  foiind  to  have  the  best  myceliimi  growth  and  therefore 
was  used  for  cultivation  and  production  of  mycorrhizae. 

Analyses  of  sterilized  peat,  on  one  hand,  and  of  peat 
inoculated  with  mycelium,  on  the  other,  have  shown  that 
fungi  of  the  species  Suillus  are  able  to  break  up  the  nutri- 
ents nitrogen  (N)  and  potassium  (K)  (table  1). 

For  mycorrhizae  inoculation,  peat  interlaced  with  myce- 
lium was  mixed  with  the  topsoil  of  seed  and  transplant 
beds  in  the  nursery  for  many  years.  Cultivation  of  cembran 
pine  plants  since  1957  has  resulted  in  nursery  soil  that  is 
now  full  of  mycelium,  and  therefore  inoculation  has  not 
been  necessary  for  many  years. 

CULTIVATION  OF  SEEDLINGS 

Cembran  pine  seeds  are  wintered  outdoors  in  thin,  al- 
ternate layers  of  seeds  and  sand  inside  wooden  boxes  that 
are  sheltered  with  wire  nets  to  exclude  mice.  In  this  way 
they  can  be  put  into  the  seedbeds  in  spring  without  strati- 
fication. The  seeds  are  lying  close  together  and  are  cov- 
ered with  thin  layers  of  compost  and  peat.  The  seedbeds 
must  be  shadowed  and  sheltered  against  mice  and  espe- 
cially nutcrackers,  which  sometimes  are  a  great  problem. 

Because  of  embryo  dormancy,  30  to  50  percent  of  seeds 
will  not  germinate  in  the  first  year,  but  will  in  the  follow- 
ing year  (fig.  4).  Normally,  cembran  pine  seeds  have  a 
germination  rate  of  about  60  to  70  percent;  this  means 
that  1,800  to  2,000  seedlings  per  square  meter  occupy  the 
seedbed  by  the  end  of  the  second  year. 

TRANSPLANTING 

After  3  years,  the  seedlings,  now  2  or  3  years  old,  are 
transplanted  in  rows;  in  the  transplant  beds  there  are 
about  120  plants  per  square  meter.  The  best  times  for 
transplanting  cembran  pine  seedlings  are  spring  or  sum- 
mer. Transplanting  can  also  be  done  in  autumn,  but  this 
can  cause  frost-heaving  problems  with  insufficiently 
rooted  transplants. 

Barerooted  transplants  are  4  or  5  years  old  (fig.  5)  when 
used  for  afforestation,  which  is  done  in  spring  or  autumn. 
Afforestation  done  in  the  spring  shows  better  results  and 
less  transplant  loss  than  in  autimin. 

POT-PLANT  PRODUCTION 

In  recent  years  our  tree  nursery  also  produces  pot 
plants.  These  are  cultivated  in  pots  made  of  pressed  peat, 


Table  1 — Analyzed  values  of  the  nutrients  N  and  K  of  sterilized  peat, 
inoculated  by  mycelium  of  Suillus  placidus  after  different 
growth  periods 


ALE 

Total 

Mycelium  growth 

NO3-N  NH,-N 

NH  -N 

4 

-  -  -  mg/100  g-  -  - 

Percent 

mg/100  g 

Percent 

Without  mycelium 

0.95  27 

0.18 

8 

0.01 

0-4  weeks 

1 .47  37 

.28 

12 

.01 

5-8  weeks 

1 .40  68 

.34 

37 

.05 

299 


Figure  4 — Cembran  pine  seedlings  of  two  differ- 
ent seed  provenances  at  the  beginning  of  the 
second  year's  growing  season  showing  different 
percentages  of  dormancy. 


Figure  5 — Transplants  of  cembran  pine,  5  and  4 
years  old.  At  this  size  they  are  fit  for  outplanting. 
A  speciality  of  the  "biological  cultivation  method" 
in  the  tree  nursery  "Klausboden"  is  the  use  of 
green-manuring  plants  instead  of  chemical  fertil- 
izers for  soil  regeneration.  This  can  be  seen  in 
the  right  half  of  the  photo. 


Figure  6 — High-elevation  afforestation  with 
cembran  pine  and  larch  for  protection  of  the  road 
crossing  the  Arlberg  Pass  between  Tyrol  and 
Vorarlberg.  Note  the  steel  snow  bridges  in  the 
background  used  for  stabilizing  the  snow  pack. 


which  can  be  pierced  by  the  roots  and  will  rot  in  a  few 
years.  Cembran  pine  seedlings  can  be  transplanted  into 
these  pots  in  spring  and  in  siunmer.  They  are  ready  for 
afforestation  after  2  or  3  months  when  their  roots  are  be- 
ginning to  pierce  the  pots.  Tests  have  shown  that  a  mix- 
ture of  90  percent  peat  and  10  percent  bruised  grape 
seeds  is  best  for  mycorrhizae  growth  in  these  pots. 

FINAL  REMARKS 

The  tree  nursery  "Klausboden"  produces  about  60,000 
transplants,  100,000  seedlings,  and  40,000  pot  plamts  of 
the  stone-pine  species  cembran  pine  each  year  on  a  bed 
area  of  0.54  hectares.  For  more  than  30  years  these 
plants  have  proved  to  be  well  adapted  to  the  harsh  envi- 
ronment of  subalpine  and  timberline  areas  in  the  Inner 
Alps  of  Austria  (fig.  6). 

Research  work,  especially  on  cembran  pine  mycorrhi- 
zae, has  been  an  important  part  of  the  research  program 


300 


on  subalpine  forests.  This  research  started  in  Austria 
after  the  avalanche  catastrophe  of  1951,  and  it  is  still  go- 
ing on.  Very  little  is  known  about  the  coevolution  of  myc- 
orrhizal  fungi  and  the  other  subalpine  stone-pine  species 
in  Asia  and  North  America. 

Because  we  can  be  sure  that  mycorrhizae  are  essential 
for  the  growth  and  survival  of  all  timberline  tree  species, 
it  is  very  important  that  intensive  research  be  started  all 
over  the  world  on  these  interrelations. 

ACKNOWLEDGMENTS 

The  author,  responsible  for  running  the  tree  nursery 
"Klausboden,"  gives  special  thanks  to  Dr.  F.  Gobi,  mycor- 
rhizae specialist,  for  her  cooperation  and  very  helpful 
ideas  concerning  the  nursery  and  this  paper. 

REFERENCES 

Czell,  A.;  Redlich,  G.  C.  1966.  Die  Beeinflussung  des 
Gebrauchswertes  von  Junglarchen  durch  kombinierte. 
Wurzel-Grimdungung  (KWGD).  Centralblatt  fur  das 
gesamte  Forstwesen,  83.,  2,  Osterr.  Agrarverlag,  Wien: 
65-84. 


Gobi,  F.  1965.  Die  Zirbenmykorrhiza  im  subalpinen 
Aufforstungsgebiet.  Centralblatt  fur  das  gesamte 
Forstwesen,  82,  Osterr.  Agrarverlag,  Wien:  89-100. 

Gobi,  F.  1979.  Erfahrimg  bei  der  Anwendung  von 
Mykorrhiza-Impfmaterial  (.Zirbe).  Centralblatt  fur 
das  gesamte  Forstwesen,  96.1,  Osterr.  Agrarverlag, 
Wien:  30-43. 

Heumader,  J.;  Gobi,  F.  1990.  Der  Forstgarten  "Klaus- 
boden". Biologische  Anzucht  von  Pflanzen  fur  Hochlagen 
seit  mehr  als  35  Jahren.  Wildbach-xmd  Lawinenverbau, 
Zeitschrift  des  Vereins  der  Diplomingenieure  der 
Wildbach-xmd  Lawinenverbautmg  Osterreichs,  Salzburg 
54,  113:  85-106. 

Leys,  E.  1970.  Erfolgreiche  Zirbenanzucht.  Gebietsbaulei- 
tung  Oberes  Inntal,  Imst.  8  p.  Unpublished  manuscript. 

Moser,  M.  1958.  Die  kvmstliche  Mykorrhizaimpfung  von 
Forstpflanzen,  11.  Die  Torfstreukultur.  Forstwissen- 
schaftl.  Centralblatt,  77,  Osterr.  Agrarverlag,  Wien: 
273-278. 

Moser,  M.  1959.  Die  kunstliche  Mykorrhizaimpfung  von 
Forstpflanzen.  III.  Die  Impfmethodik.  Forstwissen- 
schaftl.  Centralblatt,  78,  Osterr.  Agrgirverlag,  Wien: 
193-202. 


301 


WHITEBARK  PINE  CONSERVATION  IN 
NORTH  AMERICAN  NATIONAL  PARKS 

Katherine  C.  Kendall 


Abstract — Whitebark  pine  {Pinus  albicaulis)  is  a  prominent 
component  of  high-elevation  forests  in  National  Parks  in  the 
Western  United  States  and  Canada.  Park  management  attempts 
to  conserve  genetic  integrity  and  allow  natural  processes  to  occur 
unimpeded  in  these  communities.  However,  whitebark  pine  for- 
ests are  threatened  by  introduced  disease  and  fire  suppression  in 
many  areas.  Retention  of  fire-dependent,  mixed-species  white- 
bark pine  communities  will  require  aggressive  programs  to  intro- 
duce fire,  or  mimic  its  effects.  Emerging  efforts  to  develop  blister 
rust-resistant  whitebark  pine  raise  questions  about  when  and  to 
what  extent  it  is  appropriate  to  alter  park  genetic  stock.  Prece- 
dents set  by  related  cases  and  Park  Service  policy  guiding  genetic 
resource  protection  and  native  plant  restoration  are  summarized. 
A  survey  of  the  status  of  whitebark  pine  throughout  its  range 
and  studies  to  establish  genetic  and  adaptive  variation  are  most 
needed.  For  populations  at  risk,  seed  bank  collections  are  urged. 
Trade-offs  of  various  whitebark  pine  management  alternatives 
are  evaluated. 


Whitebark  pine  (Pinus  albicaulis)  is  a  component  of 
high-elevation  forests  in  15  National  Park  areas  in  the 
Western  United  States  and  Canada  (fig.  1).  Its  impact 
can  be  found  at  landscape,  community,  and  species  scales. 
Whitebark  pine  stands  influence  snow  accumulation  and 
retention,  thus  affecting  hydrological  characteristics  in 
the  drainages  where  they  occur.  As  a  pioneer  at  harsh, 
exposed  sites,  whitebark  pine  modifies  the  microclimate 
and  allows  other  vegetation  such  as  subalpine  fir  (Abies 
lasiocarpa)  to  establish  and  persist  (Habeck  1969).  A  va- 
riety of  wildlife,  including  the  threatened  grizzly  bear 
(Ursus  arctos),  rely  on  whitebark  pine  communities  for 
food  and  shelter  (Hutchins  and  Lanner  1982;  Kendall 
1983;  Mattson  and  Reinhart,  these  proceedings;  Tomback 
1978). 

United  States  and  Canadian  National  Park  policy  is  to 
preserve  natural  ecological  and  evolutionary  processes  in 
their  natural  areas.  Management  of  individual  species 
such  as  whitebark  pine  is  t3T)ically  limited  to  inventory 
and  monitoring  the  status  of  the  species.  Under  current 
U.S.  National  Park  Service  (NPS)  management  philoso- 
phy, changes  caused  by  natural  events  such  as  native 
insect  outbreaks  or  native  diseases  are  not  normally  in- 
terfered with.  Practices  to  increase  seed  production,  arti- 
ficially reforest  naturally  disturbed  stands,  or  other  man- 
agement intervention  generally  are  not  appropriate. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Katherine  C.  Kendall  is  Research  Ecologist  at  Glacier  National  Park, 
Science  Center,  West  Glacier,  MT  59936. 


Forestry  Canada  and  USDA  Forest  Service  personnel 
manage  the  bulk  of  whitebark  pine  forests.  Their  nonwil- 
demess  lands  are  managed  for  multiple  use,  and  a 
broader  range  of  activities  are  allowed  in  their  jurisdic- 
tions than  in  National  Parks.  Because  of  this,  National 
Parks  and  wilderness  areas  serve  as  benchmarks  or  ex- 
perimental controls  for  national  and  private  forests  and 
lands  managed  by  other  agencies  for  a  variety  of  pur- 
poses. Management  intervention  in  National  Parks,  how- 
ever, may  be  appropriate  when  park  resources  are 
changed  by  himian  activities. 

MANAGEMENT  CHALLENGES 

Whitebark  pine  is  in  jeopardy  in  many  areas  fi'om  intro- 
duced disease  and  fire  suppression.  The  most  serious 


X    isolated  occurrence 


Figure  1 — Whitebark  pine  occurrence  in  western 
Canadian  and  U.S.  National  Parl<  areas.  Distribution 
data  are  based  on  Arno  and  Hoff  (1 986)  and  Ogilvie 
(1990). 


302 


threat  comes  from  a  normative  disease,  white  pine  bhster 
rust  {Cronartium  ribicola),  which  cannot  be  eradicated  or 
controlled.  Whitebark  pine  mortality  exceeds  90  percent 
in  much  of  the  northwestern  portion  of  its  range  (Kendall 
and  Arno  1990).  Mortality  rates  are  lower  in  the  drier  cli- 
mates to  the  south  but  may  increase  in  the  future.  Al- 
though blister  rust  infected  most  five-needled  pine  popu- 
lations by  the  1930's,  mortality  has  not  stabilized.  Blister 
rust  is,  in  part,  responsible  for  a  rapid  decline  of  white- 
bark  pine  in  the  Bob  Marshall  Wilderness  Complex,  MT, 
in  the  last  20  years  (Keane  and  Morgan,  these  proceed- 
ings). Blister  rust  has  been  epidemic  in  sugar  pine  (P. 
lamhertiana)  in  the  southern  Sierra  Nevada  Mountains, 
CA,  since  the  1970's  where  it  had  caused  at  least  25  per- 
cent mortality  by  1985  (Kinloch  and  Dulitz  1990).  Al- 
though the  progress  of  the  white  pine  blister  rust  epi- 
demic is  not  fully  understood,  whitebark  pine  infection 
and  mortality  rates  may  increase  if  weather  patterns 
change  or  if  blister  rust  mutates  to  forms  better  able 
to  survive  dry  climates. 

The  alarming  loss  of  whitebark  pine  trees  has  broad 
repercussions  for  community  and  landscape  processes. 
As  whitebark  pine  declines,  not  only  is  mast  for  wildlife 
diminished  (Kendall  and  Arno  1990)  but  vegetative  pat- 
terns change.  In  areas  of  high  whitebark  pine  mortality, 
lack  of  shelter  was,  in  part,  responsible  for  declines  in 
associated  subalpine  fir  in  some  areas  (Keane  1992).  Pre- 
dicted changes  in  former  whitebark  pine  commimities  in- 
clude the  absence  of  reforestation  after  disturbance  in 
high-elevation,  rocky,  and  windblown  sites  and  lowered 


tree-line  elevations.  Finally,  the  hydrology  of  large  areas 
will  be  altered  as  snow  accumulation  changes  with 
vegetation. 

Natural  resistance  of  the  original  whitebark  pine  popu- 
lation to  blister  rust  is  extremely  low.  While  there  are  no 
direct  data  on  the  resistance  of  whitebark  pine,  only  one 
western  white  pine  {P.  monticola)  tree  in  10,000  is  rust 
resistant  (Bingham  1983;  Hoff  and  others,  these  proceed- 
ings). Since  whitebark  pine  is  more  susceptible  to  blister 
rust  than  western  white  pine,  the  number  of  resistant 
whitebark  pine  trees  is  likewise  less  than  0.0001  percent. 
While  blister  rust  is  not  expected  to  extirpate  whitebark 
pine  rangewide  or  even  at  most  high-hazard  sites  because 
of  the  presence  of  resistance  (Hoff  and  others  1980),  iso- 
lated populations  may  become  extinct  where  mountain 
pine  beetle  {Dendroctonus  ponderosae)  kill  the  remaining 
resistant  whitebark  pine.  Using  techniques  developed  in 
the  successful  program  to  breed  rust-resistant  western 
white  pine,  the  USDA  Forest  Service  is  investigating  the 
feasibility  of  developing  rust-resistant  whitebark  pine. 
Based  on  their  experience  with  western  white  pine,  a  pro- 
gram to  develop  resistant  whitebark  pine  is  expected  to  be 
successful  (Hoff  1991). 

Because  whitebark  pine  is  fire  dependent  in  many  ar- 
eas, fire  suppression  poses  another  threat.  Prior  to  hu- 
man intervention,  the  average  natural  fire  intervals  in 
whitebark  pine  stands  were  50  to  350  years  (Arno  and 
Weaver  1990).  Despite  natural  fire  polices,  a  serai  white- 
bark pine  stand  would  now  burn  once  every  3,000  years 
(Arno  1986).  In  mixed-species  stands  where  it  is  most 


Table  1 — Distribution  and  status  of  whitebark  pine  (PIAL^)  in  North  American  National  Park  areas 


Vegetation  Blister  Rust?  RIAL 

Park  area  map?  Area  with  RIAL  iAbsent/Z^esent/iAiknown  status 


United  States 


Big  Hole  NB 

N 

PIAL  rare 

A2 

unknown 

Crater  Lake  NP 

Y 

unknown 

P 

>45%  mortality^ 

Glacier  NP 

Y 

unknown 

P 

>90%  mortality^ 

Grand  Teton  NP 

Y 

unknown 

U  PIAL,  PIFL 

low  mortality 

Lassen  Volcanic  NP 

Y 

unknown 

P  PILA,  PIMO;  U  PIAL 

low  mortality 

Mount  Rainier  NP 

Y 

unknown 

P  PIMO,  PIAL 

Blister  rust/mature  PIAL: 

92%  infection;  43%  dead^ 

North  Cascades  NP 

Y 

3,922  ha« 

U 

unknown 

Olympic  NP 

N 

unknown 

P  PIMO;  U  PIAL 

poor  reproduction 

Sequoia-Kings  Canyon  NP 

Y 

unknown 

P  PIBA,  PILA,  PIMO 

healthy 

Yellowstone  NP 

Y 

88,500  ha^ 

U  PIAL.  PIFL 

low  mortality  PIAL 

Yosemite  NP 

Y 

12,727  ha« 

P  PILA;  A^  PIAL,  PIMO 

healthy 

anada 

Banff  NP 

Y 

300  ha 

U 

unknown 

Jasper  NP 

Y 

57,730  ha 

U 

no  apparent  threats  . 

Kootenay  NP 

Y 

unknown 

U 

unknown 

Waterton  Lakes  NP 

Y 

3,990  ha 

P  PIAL,  PIFL 

serious  mortality; 

extent  unknown 


'PIAL  =  Pinus  albicaulis;  PIBA  =  P.  balfouriana;  PIFL  =  P.  flexilis;  PILA  =  P.  lambertiana;  PIMO  =  P.  monticola. 
^Absence  assumed  but  no  surveys  conducted. 
Trom  Jackson  and  Faller  (1973). 
^From  Kendall  and  Arno  (1990). 

^Figures  from  survey  of  Sunrise  Ridge  Trail  on  August  10, 1992  (Hoff  1992). 
®From  Agee  and  others  (1985). 
Trom  Renkin  and  Despain  (1992). 
Trom  Jan  Van  Wagtendonk  ( 1 992) . 


303 


valuable  to  bears  and  red  squirrels  (Tamiasciurus  hud- 
sonicus),  whitebark  pine  is  being  replaced  by  shade-toler- 
ant conifers  in  the  absence  of  fire.  Lack  of  fire  has  also  in- 
creased whitebark  pine  mortality  from  mountain  pine 
beetle  and  dwarf  mistletoe  {Arceuthohium  spp.)  (Kendall 
and  Arno  1990).  Fire  suppression  creates  even-aged 
lodgepole  pine  (P.  contorta)  stands,  which  can  support 
massive  buildups  of  pine  beetles  that  spill  over  into  adja- 
cent whitebark  pine  forests.   Fire  is  a  primary  regulator 
of  the  frequency  and  intensity  of  dwarf  mistletoe  infec- 
tion. Efforts  to  enhance  the  resistance  of  whitebark  pine 
to  blister  rust  should  go  hand-in-hand  with  ensuring  con- 
ditions that  maintain  whitebark  pine  on  the  landscape. 
Thus  far,  despite  widespread  natural  fire  policies,  the 
political  constraints  on  prescribed  fire  mean  most  are 
controlled. 

Park  Service  ability  to  deal  with  these  threats  is  limited 
by  how  little  we  know  of  whitebark  pine  communities.  Of 
15  parks  with  whitebark  pine,  most  had  basic  vegetation 
maps  but  only  half  knew  the  amount  of  area  covered  by 
whitebark  pine  (table  1).  Where  available,  the  whitebark 
pine  coverage  figures  are  not  necessarily  comparable. 
Some  estimates  only  included  areas  where  whitebark  pine 
was  dominant  or  codominant,  while  others  included  the 
area  of  all  communities  with  whitebark  pine  present. 

Because  none  of  the  parks  had  actually  surveyed  white- 
bark pine,  its  status  in  most  areas  was  based  on  casual,  at 
times  out-dated,  observation  or  was  unknown.  For  ex- 
ample, the  whitebark  pine  mortality  estimate  for  Glacier 
National  Park  (>90  percent;  table  1)  was  based  on  obser- 
vations made  in  the  early  1970's  during  a  search  for  blis- 
ter rust-resistant  trees.  Rocky  Mountain  National  Park, 
CO,  lists  whitebark  pine  as  present,  but  it  is  unlikely  to 
occur  so  far  south  of  its  range  and  was  probably  confused 
with  limber  pine  (P.  flexilis).  Although  blister  rust  was 
verified  as  present  in  eight  parks,  because  most  park 
staffs  were  unfamiliar  with  blister  rust,  its  presence  was 
unknown  or  assumed  absent  in  seven  parks.  Few  parks 
had  information  on  the  effects  of  fire  suppression  in  their 
whitebark  pine  commimities. 

Whitebark  pine  forests  in  National  Parks  have  received 
little  research  attention.  Canadian  Parks  Service  re- 
ported no  ongoing  research  in  whitebark  pine  communi- 
ties. In  the  United  States,  only  Yellowstone  National 
Park  is  currently  conducting  whitebark  pine  research. 
Some  studies  on  postfire  succession  and  fire  effects  on  soil 
seed  banks  in  Yellowstone  National  Park  include  white- 
bark pine  stands.  Yellowstone  National  Park  personnel 
have  monitored  cone  production  in  whitebark  pine  stands 
since  1981  and  will  survey  those  trees  for  blister  rust  this 
year.  No  other  parks  are  conducting  or  sponsoring  re- 
search or  monitoring  of  whitebark  pine. 

RESTORATION  POLICY 

The  vast  losses  of  whitebark  pine  and  the  possibility 
that  selectively  bred  whitebark  pine  will  be  available  in 
the  future  raise  some  interesting  questions  for  park  man- 
agers. One  challenge  will  be  to  determine  what  action  is 
appropriate  under  these  circumstances.  Canadian  Parks 
Service  policy  prescribes  active  intervention  when 
"...natural  processes  have  been  altered  by  man  and 


manipulation  is  required  to  restore  the  natm-al  balance" 
(Parks  Canada  1979).  NPS  policy  (National  Park  Service 
1988)  explicitly  provides  for  active  management  to  reverse 
the  loss  of  whitebark  pine.  Policy  states  that  the  Service 
"...will  strive  to  protect  the  full  range  of  genetic  types 
(genotypes)  native  to  plant  and  animal  populations  in  the 
parks  by  perpetuating  natural  evolutionary  processes  and 
minimizing  human  interference  with  evolving  genetic  di- 
versity." The  National  Park  Service  will  strive  to  restore 
native  plants  if  "the  species  was  substantially  diminished 
as  a  result... of  human-induced  change...." 

Park  use  of  planting  stock  with  enhanced  resistance  to 
rust  is  also  clearly  appropriate.  Policy  specifies  "where 
a  natural  area  has  become  so  degraded  that  restoration 
with  native  species  has  proven  unsuccessful,  improved 
varieties  or  similar  native  species  may  be  used"  (National 
Park  Service  1988).  For  park  restoration  projects,  resis- 
tant stock  development  should  be  sensitive  to  maintaining 
pristine  gene  pools.  "Whenever  possible,  revegetation  ef- 
forts in  natural  zones  will  use  seeds,  cuttings,  or  trans- 
plants representing  species  and  gene  pools  native  to  the 
ecological  portion  of  the  park  in  which  the  restoration 
project  is  occurring"  (National  Park  Service  1988).  Gen- 
eral guidelines  to  prevent  genetic  contamination  during 
vegetation  restoration  projects  are  beginning  to  emerge, 
but  more  specific  guidelines  await  species-specific  genetic 
research  (Potter  and  Kurth  1992). 

In  other  North  American  natural  areas,  introduced 
disease  pests  have  devastated  native  forest  trees  such  as 
American  chestnut  (Castanea  dentata),  flowering  dogwood 
(Cornus  florida),  and  butternut  {Juglans  cinerea).  The  ex- 
otic fungus,  Chinese  chestnut  blight  {Endothia  para- 
sitica), has  killed  all  mature  American  chestnut  trees 
(present  on  approximately  9  million  acres)  in  North 
America  since  its  introduction  around  1900  (Langdon  and 
Johnson  1992).  Early  attempts  to  control  the  disease  and 
develop  a  resistant  hybrid  of  Oriental  and  American 
chestnuts  were  not  successful.  Current  efforts  are  back- 
crossing  hybrids  to  yield  stock  with  few  "Oriental"  traits, 
but  it  remains  to  be  seen  if  the  stock  is  resistant.  Dog- 
wood anthracnose  (Discula  destructiva),  first  found  in 
North  America  in  the  1970's,  can  kill  all  flowering  dog- 
wood (Cornus  florida)  at  sites  favorable  to  this  introduced 
fungus.  Another  exotic  fungus,  butternut  canker  (Serio- 
coccus  clauigineti-juglandacearum),  caused  80  percent  de- 
clines in  butternut  in  North  and  South  Carolina  since  the 
early  1970's  (Anderson  1990)  and  suppressed  nut  crops  in 
Great  Smoky  Mountains  National  Park  (Langdon  and 
Johnson  1992).  Apparently  resistant  flowering  dogwood 
from  Catoctin  Mountain  Park  and  putatively  resistant 
butternut  from  throughout  its  range  are  currently  under- 
going screening  to  verify  resistance.  Resistant  stock 
would  be  used  for  introduction  to  natural  populations  and 
for  further  breeding  work  (Langdon  and  Johnson  1992). 

National  Park  Service  policy  guiding  genetic  conserva- 
tion is  rapidly  evolving.  Langdon  and  Johnson  (1992)  felt 
that  the  appropriateness  of  using  hybrids  (such  as  those 
being  developed  for  the  American  chestnut)  for  restora- 
tion of  natural  zones  had  not  been  established.  However, 
a  report  clarifying  Park  Service  policy  (Keystone  Center 
1991)  stated,  "Restoration  of  native  species  is  encouraged 
where... the  restored  species  most  nearly  approximates 


304 


the  extirpated  species,  and  the  species  disappeared 
because  of  hiunan-induced  impacts  to  the  population 
or  ecosystem.  Exotic  species  may  not  be  introduced  into 
natural  zones  of  parks  except  where  they  are  the  nearest 
living  relatives  of  extirpated  native  species...." 

Use  of  rust -resistant  whitebark  pine  stock  for  restora- 
tion in  natural  park  zones  is  more  straightforward.  The 
efforts  to  develop  resistant  whitebark  pine  focus  on  speed- 
ing natural  selection  processes  and  do  not  involve  hybrid- 
ization or  genetic  engineering.  In  the  face  of  the  devastat- 
ing loss  of  whitebark  pine  in  many  areas,  use  of  these 
materials  clearly  would  be  in  hne  with  National  Park  Ser- 
vice policy.  Moreover,  to  ensure  that  park  genotypes  are 
preserved  in  the  breeding  program  it  would  be  in  the  Park 
Service's  best  interests  to  be  actively  involved  with  the 
rust-resistance  program. 

However,  because  natural  selection  will  no  doubt  differ 
from  our  best  attempts  to  mimic  it  (in  the  breeding  pro- 
gram), I  believe  there  is  also  value  in  having  areas  where 
whitebark  pine  communities  are  left  undisturbed  to  re- 
spond to  the  changes  wrought  by  himians.  Practically 
speaking,  this  will  be  easy  to  achieve  since  it  will  be  logis- 
tically  impossible  to  intervene  in  much  of  whitebark  pine's 
range. 

MANAGEMENT  OPTIONS 

Long-Hved  and  slow-growing,  whitebark  pine  trees  are 
50  to  100  years  old  before  they  begin  to  produce  signifi- 
cant cone  crops.  Thus,  whitebark  pine  stands  are  espe- 
cially slow  to  recover  from  damage  and  slow  to  respond 
to  management  measures  (Kendall  and  Amo  1990).  Until 
fire  is  allowed  to  play  its  historical  role  in  whitebark  pine 
habitats  and  a  high  level  of  resistance  to  blister  rust  is 
achieved,  whitebark  pine  will  continue  to  decline  or  per- 
sist in  many  areas  at  very  low  levels. 

We  can  expect  to  lose  (or  may  have  already  lost)  some 
small,  isolated  populations.  In  areas  of  high  mortality, 
svirviving  trees  represent  only  a  subset  of  the  original  ge- 
netic pool.  Park  Service  action  strategy  must  recognize 
whitebark  pine's  special  attributes  and  should  be  based 
on  evaluation  of  the  trade-offs  of  various  management  al- 
ternatives. A  viable  program  should  contain  a  combina- 
tion of  the  following  options  and  recommendations. 

Restoration  and  Revegetation 

Management  options  for  general  restoration  of 
whitebark  pine  are: 

1.  Where  fire  suppression  has  caused  the  decline  of 
whitebark  pine,  assess  historical  and  present  whitebark 
pine  forest  composition.  Prepare  sites  for  regeneration 
with  management-initiated  burns  or  manual  removal  of 
competing  species.  In  natural  fire  zones,  ease  prescrip- 
tions so  that  natiiral  fires  will  be  allowed  to  occvir  in 
whitebark  pine  communities. 

2.  Rely  on  natural  regeneration.  Regeneration  may  fail 
to  occur  or  may  be  very  slow  if  seed  sources  are  beyond 
the  range  of  Clark's  nutcrackers  {Nucifraga  columhiana) 
or  are  rare  and  scant. 

3.  Replant  seed  or  seedlings  of  wild  or  resistant  stock. 


The  factors  affecting  whitebark  pine  establishment  are 
only  beginning  to  be  studied  (McCaughey  1990).  If  suc- 
cessful, planting  will  speed  reestablishment  of  whitebark 
pine.  Planting  resistant  varieties  on  highly  productive 
sites  will  accelerate  growth  and  presumedly  result  in 
early  cone  production. 

Rust  Resistance 

Management  options  for  dealing  with  rust-resistance  is- 
sues are  (Hoff  1991;  Hoff  and  others,  these  proceedings): 

1.  No  intervention.  Allow  natxu*al  processes  to  develop 
rust-resistant  whitebark  pine  and  restore  whitebark  pine 
communities.  This  may  be  a  better  selector  of  the  various 
resistance  mechanisms,  but  will  require  a  great  amoiint  of 
time  given  whitebark  pine's  slow  growth  and  maturation. 
Vegetation  and  animal  communities  will  be  dramatically 
altered  for  an  extremely  long  time.  This  option  may 
retain  more  genetic  diversity  than  would  a  breeding 
program;  however,  genetic  variation  will  decline  if  some 
small  populations  are  lost  or  are  greatly  diminished. 

2.  Propagate  and  plant  stock  from  surviving  trees 
foimd  in  areas  of  high  blister  rust  mortality.  These  mate- 
rials are  available  now  and  any  resistance  present  is  a  re- 
sult of  natural  selection.  This  would  mesh  well  with  a 
seed  bank  program. 

3.  Manage  natural-selection  stands.  Inhighbhster 
rust  hazard  areas  with  high  whitebark  pine  seedling  den- 
sity, allow  natiiral  selection  to  act  on  naturally  regenerat- 
ing stock.  Shorten  generation  times  by  removing  compet- 
ing species.  The  advantages  and  disadvantages  discussed 
for  the  no-intervention  option  hold  true  here.  Another 
beneficial  aspect  is  that  the  stock  or  seed  for  outplanting 
will  be  adapted  to  local  conditions.  It  is  debatable 
whether  clearing  competing  species  is  appropriate  in 
National  Parks.  However,  if  parks  do  not  implement  this 
option,  park  genetic  stock  vnll  not  be  part  of  this  impor- 
tant component  of  any  plan  developed  for  whitebark  pine 
conservation. 

4.  Breeding  program.  One  option  is  to  collect  wind-pol- 
linated seed  or  cross  pollinate  trees  in  situ  that  exhibit 
some  resistance  to  blister  rust,  germinate  resultant  seed, 
and  infect  seedlings  with  rust  to  select  resistant  tj^Des. 
The  surviving  seedlings  would  be  outplanted,  but  some 
would  be  used  to  establish  seed  orchards.  Repeat  for  sev- 
eral generations  for  resistant  seed  stock.  Another  option 
is  to  collect  scion  wood  from  resistant  trees,  graft  to  nurs- 
ery root  stock,  test  for  resistance  and  induce  flowering, 
and  cross  with  other  resistant  types.  Although  these  pro- 
grams may  be  less  successful  in  selecting  for  resistance 
than  nature  woiild  be,  they  vdll  generally  reduce  the  time 
required  to  develop  a  high  degree  of  rust  resistance 
(65-100  years  vs.  himdreds  of  years).  A  breeding  program 
could  be  designed  to  preserve  genotypes  from  any  tar- 
geted area  and  thus  preserve  representatives  from  popu- 
lations at  risk  of  being  lost. 

RECOMMENDATIONS 

A  nimiber  of  recommendations  for  the  management  of 
whitebark  pine  on  National  Park  Service  lands  can  be 
made  at  this  time. 


305 


1.  Inventories  and  monitoring  are  recognized  needs  not 
yet  fully  realized  by  National  Parks.  Most  National  Parks 
with  whitebark  pine  have  not  surveyed  the  extent  and 
status  of  whitebark  communities  (table  1).  We  need  to 
learn  how  much  whitebark  pine  now  exists  and  how  much 
has  been  lost  to  various  causes.  Resistant  trees  should  be 
located  and  marked. 

2.  Research  genetic  and  adaptive  variation.  Because 
whitebark  pine  is  wind-pollinated  and  seed  is  cached  by 
Clark's  nutcrackers  up  to  23  km  from  the  seed  soiirce 
(Vander  Wall  and  Balda  1977),  total  genetic  variation  in 
whitebark  pine  populations  may  be  relatively  high  and,  be- 
cause of  this  constant  genetic  mixing,  patterns  of  adaptive 
genetic  variation  may  be  relatively  broad.  That  not  with- 
standing, genetic  variation  should  be  established  to  pro- 
vide guidance  for  park  restoration  programs. 

3.  In  areas  where  it  is  unknown,  determine  the  fire  his- 
tory of  whitebark  pine  stands.  Study  how  to  reintroduce 
fire  to  whitebark  pine  habitats  or  how  to  mimic  its  effects 
in  the  variety  of  habitats  where  whitebark  pine  occurs. 
Apply  results  to  an  experimental  management  program. 

4.  Support  research  on  white  pine  blister  rust  damage 
and  mortality  rate  equilibrium  points  for  whitebark  pine 
populations  throughout  its  range. 

5.  Establish  managed  natural-selection  stands  for  all 
whitebark  pine  populations  at  risk. 

6.  Support  and  collaborate  with  on-going  efforts  to  de- 
velop rust-resistant  stock.  Develop  guidelines  for  select- 
ing genetic  stock  for  inclusion  in  a  breeding  program  that 
will  meet  Park  Service  mandates. 

7.  Seed  banks.  Collect  and  store  seed  to  represent  the 
breadth  of  genotypic,  phenotypic,  and  geographic  varia- 
tion. Collection  from  small,  isolated  populations  should 
be  a  priority.  Seed  stored  under  optimum  conditions  can 
remain  viable  for  30  to  100  years  or  more  and  represent 
insurance  against  catastrophic  loss  in  natural  settings. 

8.  Develop  a  comprehensive  strategy  for  the  conserva- 
tion of  whitebark  pine  in  cooperation  with  other  land 
management  agencies  and  conservation  organizations. 
One  component  should  include  the  end  of  practices  that 
remove  whitebark  regeneration  in  clearcuts  and  burns. 
The  reverse,  eliminating  whitebark  pine's  competitors, 
should  be  adopted. 

CONCLUSIONS 

While  whitebark  pine  is  not  alone  in  its  need  of  atten- 
tion in  our  quest  to  preserve  biodiversity,  it  is  under  siege 
and  we  know  little  about  it.  High-mountain  ecosystems 
are  particularly  vulnerable  to  disturbance.  Thus,  white- 
bark pine  will  be  more  susceptible  than  most  species  to 
the  effects  of  global  climate  change,  acid  precipitation, 
and  other  forms  of  air  pollution  and  should  serve  as  a 
sensitive  monitor  of  subtle  environmental  disturbance. 
Whitebark  pine  research  and  conservation  will  require 
coordinated  action  between  various  agencies  and  mem- 
bers of  the  conservation  community.  I  urge  agencies  to 
conduct  basic  inventories  of  our  subalpine  environments 
and  create  a  comprehensive  conservation  strategy  for 
whitebark  pine  before  more  genetic  material  is  lost.  Be- 
cause of  their  preservation  missions.  National  Parks  must 
be  key  players  in  conserving  this  as  well  as  other  species 
at  risk. 


ACKNOWLEDGMENTS 

I  am  indebted  to  R.  Hoff  for  extensive,  invaluable  dis- 
cussions on  whitebark  pine  genetics  and  management  op- 
tions for  the  development  of  rust  resistance.  Thank  you  to 
S.  Woodley  for  information  on  the  status  of  whitebark 
pine  in  Canadian  National  Parks.  R.  Keane,  L.  Kurth, 
W.  McCaughey,  and  W.  Schmidt  provided  valuable  com- 
ments on  an  earlier  draft. 

REFERENCES 

Agee,  J.  K.;  Pickford,  S.  G.;  Kertis,  J.;  Finney,  M.; 
de  Gouvenain,  R.;  Quinsey,  S.;  Nyquist,  M.;  Root,  R.; 
Stitt,  S.;  Waggoner,  G.;  Titlow,  B.  1985.  Vegetation  and 
fuel  mapping  of  North  Cascades  National  Park  Service 
complex.  Seattle,  WA:  National  Park  Service  Coopera- 
tive Park  Studies  Unit,  College  of  Forest  Resources, 
University  of  Washington:  17,59. 

Arno,  Stephen  F.  1986.  Whitebark  pine  cone  crops;  a  di- 
minishing source  of  wildlife  food?  Western  Journal  of 
AppHed  Forestry.  1(3):  92-94. 

Arno,  Stephen  A.;  Hoff,  Raymond  J.  1989.  Silvics  of 
whitebark  pine  (Pinus  albicaulis).  Gen.  Tech.  Rep.  INT- 
253.  Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station,  lip. 

Arno,  Stephen  A.;  Weaver,  Tad.  1990.  Whitebark  pine 
community  types  and  their  patterns  on  the  landscape. 
In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
997-105. 

Bingham,  R.  T.  1983.  Blister  rust  resistant  western  white 
pine  for  the  Inland  Empire:  the  story  of  the  first  25 
years  of  the  research  and  development  program.  Gen. 
Tech.  Rep.  INT-146.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermountain  Forest  and 
Range  Experiment  Station.  45  p. 

Habeck,  James  R.  1969.  A  gradient  analysis  of  a  timber- 
line  zone  at  Logan  Pass,  Glacier  Park,  Montana.  North- 
west Science.  43(2):  65-73. 

Hoff,  R.  J.;  Bingham,  R.  T.;  McDonald,  G.  I.  1980.  Relative 
blister  rust  resistance  of  white  pines.  European  Journal 
of  Forest  Pathology.  10:  307-316. 

Hoff,  Ray.  1991.  [Personal  communication].  October  10. 
Moscow,  ID:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station. 

Hoff,  Ray.  1992.  [Personal  communication].  August  17. 
Moscow,  ID:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Jackson,  M.  T.;  Faller,  A.  1973.  Structural  analysis  and 
dynamics  of  the  plant  communities  of  Wizard  Island, 
Crater  Lake  National  Park.  Ecological  Monographs.  43: 
441-461. 

Keane,  R.  T.  1992.  [Personal  communication].  December  8. 
Data  on  file  at  U.S.  Department  of  Agriculture,  Forest 


306 


Service,  Intermountain  Research  Station,  Intermoun- 
tain  Fire  Sciences  Laboratory,  Missotila,  MT. 
Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  grizzly  and  black 
bears  in  the  Yellowstone  area.  In:  International  Confer- 
ence on  Bear  Research  and  Management:  Proceedings. 
Madison,  WI:  International  Bear  Association.  5: 
166-173. 

Kendall,  K.  C;  Arno,  S.  F.  1990.  Whitebark  pine— an  im- 
portant but  endangered  wildlife  resource.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposiimi  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989  March 
29-30;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermoxmtain  Research  Station:  264-274. 

Keystone  Center.  1991.  Final  consensus  report  of  the  Key- 
stone policy  dialogue  on  biological  diversity  on  Federal 
lands.  Keystone,  CO:  The  Keystone  Center.  96  p. 

Kinloch,  B.  B.;  DiUitz,  D.  1990.  White  pine  bhster  rust 
at  Mountain  Home  Demonstration  State  Forest:  a  case 
study  of  the  epidemic  and  prospects  for  genetic  control. 
Res.  Pap.  PSW-204.  Berkeley,  CA:  U.S.  Department  of 
grictdture,  Forest  Service,  Pacific  Southwest  Research 
Station.  7  p. 

Langdon,  K.  R.;  Johnson,  K.  D.  1992.  Alien  forest  insects 
and  diseases  in  eastern  USNPS  units:  impacts  and  in- 
terventions. George  Wright  Forum.  9:  2-14. 

McCaughey,  Ward  W.  1990.  Biotic  and  microsite  factors 
affecting  Pinus  alhicaulis  establishment  and  siirvival. 
Bozeman,  MT:  Montana  State  University.  78  p. 
Dissertation. 


National  Park  Service.  1988.  Management  policies.  Wash- 
ington, DC:  U.S  Department  of  the  Interior,  National 
Park  Service:  4:1-11. 

Ogilvie,  R.  T.  1990.  Distribution  and  ecology  of  whitebark 
pine  in  western  Canada.  In:  Schmidt,  W5mian  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — s5anposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-moimtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  54-60. 

Parks  Canada.  1979.  Parks  Canada  policy.  Ottawa,  ON: 
Parks  Canada,  Department  of  Indian  and  Northern 
Affairs.  68  p. 

Potter,  R.;  Kurth,  L.  1992.  Interim  genetic  guidelines  for 
restoration  projects,  Glacier  National  Park.  Glacier  Na- 
tional Park  Report.  West  Glacier,  MT:  U.S.  Department 
of  the  Interior,  National  Park  Service.  6  p. 

Renkin,  R.  A.;  Despain,  D.  G.  1992.  Fuel  moisture,  forest 
type,  and  lightning-caused  fire  in  Yellowstone  National 
Park.  Canadian  Journal  of  Forest  Research.  22(1): 
37-45. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. Living  Bird.  16(1977):  123-160. 

Van  Wagtendonk,  Jan.  1992.  [Personal  commimication]. 
July  20.  Yosemite  National  Park,  CA:  U.S.  Department 
of  the  Interior,  National  Park  Service. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of 
the  Clark's  nutcracker  and  the  pinon  pine  for  efficient 
seed  harvest  and  dispersal.  Ecological  Monographs.  47: 
89-111. 


307 


308 


RESEARCH  NEEDS  IN  WHITEBARK 
PINE  ECOSYSTEMS 


Wyman  C.  Schmidt 


Abstract — Whitebark  pine  (Pinus  albicaulis)  ecosystems  occupy 
high  elevations  over  a  large  geographic  area  in  mountains  of  the 
Western  United  States  and  Canada.  Interest  in  these  ecosystems 
has  increased  dramatically,  partly  because  whitebark  pine  for- 
ests are  very  important  for  survival  of  endangered  grizzly  bears 
(JJrsus  arctos  horribilis).  Research  has  been  sporadic.  The  ad- 
vent of  Ecosystem  Management  is  creating  new  demands  for  in- 
formation concerning  ecological  processes  and  the  need  to  man- 
age on  a  landscape  scale.  This  paper  outlines  obvious  gaps  in 
knowledge  of  this  important  species  and  ecosystem  as  expressed 
by  researchers  and  land  managers  at  a  recent  whitebark  pine 
symposium  and  workshop. 


Research  in  whitebark  pine  (Pinus  albicaulis)  ecosystems 
is  relatively  new  in  North  America.  Whitebark  pine  for- 
ests were  largely  ignored  by  managers  and  researchers  in 
the  United  States,  and  even  more  so  in  Canada.  Most  of 
this  research  has  been  done  in  the  last  10  to  15  years,  pri- 
marily in  the  Greater  Yellowstone  Ecosystem  in  Idaho, 
Wyoming,  and  Montana  (McCaughey  and  Weaver  1990). 

Important  research  has  been  under  way  in  other  areas 
vrithin  the  range  of  whitebark  pine,  but  the  overwhelming 
importance  of  this  species  in  the  Yellowstone  Ecosystem 
prompted  the  acceleration  of  research  in  that  area.  These 
forests  harbor  grizzly  bear,  moimtain  sheep,  elk,  and  a 
host  of  other  wildlife  species;  provide  late-season  water 
for  valleys  below;  provide  expansive  views  and  solitude  for 
the  high-mountain  visitor;  and,  to  a  lesser  extent,  provide 
some  of  the  wood  products  for  North  America. 

Research  in  whitebark  pine  forests  has  largely  been  done 
as  a  labor  of  love  by  individual  scientists  from  a  broad 
spectrum  of  disciplines,  not  as  a  centralized,  folly  coordi- 
nated research  program.  Researchers — from  various  uni- 
versities, the  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, U.S.  Department  of  the  Interior,  National  Park 
Service,  and  other  public  and  private  individuals — ^have 
a  professional  scientific  bond  but  few  administrative  ties. 
For  most  researchers,  whitebark  pine  studies  were  a 
small  segment  of  their  total  research  effort. 

During  most  of  the  20th  centxiry  there  has  been  a  grad- 
ual transition  from  custodial  to  more  proactive  manage- 
ment of  National  Forests.  For  most  of  the  first  half  of  this 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Wyman  C.  Schmidt  is  Project  Leader  of  the  Subalpine  Forest  Ecology 
and  Silviculture  Research  Unit,  Intermountain  Research  Station,  Forestry 
Sciences  Laboratory,  Forest  Service,  U.S.  Department  of  Agriculture, 
Montana  State  University,  Bozeman,  MT  59717-0278. 


century,  efforts  were  concentrated  on  fire,  insect,  and 
disease  protection.  Fire  protection  was  especially  efifec- 
tive,  at  least  in  the  short  term.  Custodial  management 
was  gradually  converted  to  proactive  management  with 
greater  emphases  on  production  and  regulation  of  timber, 
range,  wildlife,  and  water,  because  timber  and  range 
management  practices  emphasized  those  tangible  com- 
modities. Most  proactive  management  was  in  the  lower 
to  mid-elevation  areas;  high-elevation  site  management 
remained  largely  custodial. 

In  the  1950's  there  was  a  gradual  transition  toward 
management  that  emphasized  more  balanced  resource 
management  goals.  Recreation  became  a  significant  con- 
sideration. This  drive  for  balance  in  forest  uses  culmi- 
nated in  the  Congressionally  mandated  Multiple-Use  Act 
of  1964  for  national  forests.  This  law  continues  to  be  in 
efifect.  Many  other  public  and  private  land  managers 
have  also  adopted  the  multiple-use  concept. 

In  the  late  1980's  and  early  1990's,  the  Forest  Service 
added  ecological  criteria  for  managing  national  forests. 
These  practices,  now  termed  Ecosystem  Management, 
include  increased  emphases  on  sustaining  long-term  pro- 
ductivity, retaining  biological  diversity,  siistaining  ecologi- 
cal functions  and  integrity,  including  social  and  economic 
values  in  planning,  and  incorporating  landscape-scale  con- 
siderations into  management  practices.  A  particularly 
important  goal  is  the  continual  acquisition  and  use  of  the 
best  scientific  facts  and  concepts  dealing  with  the  various 
biological,  social,  and  economic  functions  of  the  ecosystem. 
Thus,  very  strong  partnerships  between  management,  re- 
search, education,  and  the  public  must  be  formed  and  are 
key  to  effective  implementation  of  Ecosystem  Management. 

The  initiation  of  Ecosystem  Management  led  to  realiza- 
tion that  the  scientific  base  needed  for  managing  our  vari- 
ous North  American  ecosystems  is  often  fragmented, 
incomplete,  difficult  to  locate,  not  integrated,  and  inade- 
quately interpreted.  This  was  particvdarly  apparent  in 
high-elevation  forests  where  whitebark  pine  is  a  major 
constituent. 

In  response  to  the  accelerating  demand  for  scientific  in- 
formation about  whitebark  pine  forests,  we  sponsored  a 
major  symposium  (Schmidt  and  McDonald  1990)  and  sev- 
eral workshops  were  held  to  examine  in  depth  these  high- 
moimtain  areas.  Our  objective  was  to  assemble  scientists, 
educators,  msinagers,  and  the  public  interested  in  this 
subject.  In  the  process  we  attempted  to  ferret  out  and 
present  the  best  available  information. 

The  1989  symposiimi  on  whitebark  pine  (Schmidt  and 
McDonald  1990)  closed  vnth.  a  chapter  asking  "Where  Do 
We  Go  From  Here?"  A  portion  of  that  chapter,  as  well  as 
many  of  the  individual  papers  in  the  proceedings,  dis- 
cussed knowledge  gaps  and  research  needs  in  whitebark 


309 


pine  ecosystems.  From  these  sources,  I  extracted  the 
research  needs  pointed  out  by  the  authors — mostly  re- 
searchers but  also  some  managers.  Because  research 
needs  varied  widely  in  scale,  I  stratified  them  into  differ- 
ent levels  fi'om  micro  to  macro  (gene  to  global). 

There  were  over  160  research  needs  suggested,  but  many 
were  essentially  duplicates.  I  categorized  the  suggested 
research  needs  into  nine  different  levels:  Gene,  Tree, 
Stand,  Forest,  Ecosystem,,  Landscape,  Regional,  Continen- 
tal, and  Global.  The  level  assigned  the  various  sugges- 
tions was  a  subjective  assignment  by  this  author,  but 
most  were  fairly  definitive  and  easily  categorized.  Under 
each  of  the  different  levels  I  paraphrased  the  various  re- 
search needs,  some  of  which  were  mentioned  by  one  or 
several  authors.  The  following  section  lists  those  needs. 

RESEARCH  NEEDS 

Each  heading  is  the  level  to  which  research  needs  were 
assigned.  The  statements  below  it  are  expressions  of 
what  we  need  to  know  to  fill  gaps  in  existing  knowledge 
of  whitebark  pine  ecosystems. 

Gene 

•  Genotypic  and  adaptive  variation  of  whitebark  pine. 

•  Genetic  resistance  to  blister  rust  of  whitebark  pine. 

•  Monoterpene  differences  between  geographic  areas. 

•  Effects  on  the  genetic  base  by  blister  rust  and  fire 
suppression. 

Tree 

•  How  tree  growth  dynamics  relate  to  soil  chemical 
and  physical  properties  and  geomorphology. 

•  Site,  stand,  and  area  relationships  to  mountain  pine 
beetle,  secondary  bark  beetles,  and  cone  and  seed  insects. 

•  Regeneration  mechanisms  and  requirements  of 
whitebark  pine. 

•  Techniques  for  propagating  whitebark  pine  for  a 
variety  of  site  conditions. 

•  How  to  better  identify  and  predict  cone  crop 
periodicity. 

•  Physiological  tolerances  and  ecological  characteris- 
tics, including  water  and  nutrient  relationships,  of  species 
adapted  to  major  disturbances,  such  as  mining. 

•  Characteristics  of  mycorrhizal  and  nitrogen-fixing 
symbionts  of  associated  subalpine  species. 

Stand 

•  Rehabilitation  methods  for  dealing  with  disturbances 
due  to  mining,  fire,  grazing,  and  high-use  recreation. 

•  Physiological  and  morphological  response  informa- 
tion in  relation  to  competing  tree  species. 

•  Interrelationships  of  various  resources,  such  as 
water,  forage,  cover,  and  cone  production. 

•  Successional  relationships  at  treeline  and  downward 
into  mixed  species  stands. 

•  How  to  explain  the  diversity  of  growth  forms. 


•  Role  of  domestic  animals,  such  as  cattle,  sheep,  and 
horses  in  the  introduction  of  imdesirable  plants  or  other 
organisms. 

•  Combinations  of  tree  species  and  silvictdtural  pre- 
scriptions needed  to  maintain  viable  squirrel  populations. 

•  Squirrel  population  relationships  to  cone  crops  of 
whitebark  pine. 

•  Harvesting,  thinning,  and  burning  practices  needed 
to  favor  whitebark  pine  establishment  and  growth. 

•  In  grizzly  bear  areas  where  cone  production  is  criti- 
cal, what  ecological  habitats  should  be  featured  in  re- 
search and  management. 

•  Stand  structure,  or  species  composition,  critical  to 
squirrel  behavior  and  population. 

•  Amount  of  whitebark  pine  needed  to  attract  nut- 
crackers; level  of  cache  recovery  by  nutcrackers  or  other 
predators. 

•  Relationship  of  stand  conditions,  such  as  those  cre- 
ated by  thinning  for  increasing  cone  production,  to  snow 
redistribution. 

•  Management  strategies  needed  (and  to  be  avoided) 
to  protect  and  enhance  fisheries  in  high-sdtitude  lakes 
and  streams. 

•  How  we  can  use  this  information  in  reclamation 
projects. 

Forest 

•  Better  methods  for  protecting  forests  from  insect, 
disease,  and  fire  problems. 

•  Ecological  effects  of  fire  suppression  and  how  to  rein- 
troduce fire. 

•  How  to  develop  strategies  for  protecting  rare  and  en- 
dangered species. 

•  How  fragile  or  elastic  these  forests  are  in  the  long 
term. 

•  What  conflicts  there  are  between  domestic  stock  and 
wildlife,  such  as  grizzly  bear  and  elk. 

•  Specific  squirrel/bear/whitebark  pine  relationships 
in  different  ecological  habitats. 

•  How  whitebark  pine  fits  in  the  hydrologic  cycle. 

•  How  to  set  management  objectives. 

Ecosystem 

•  How  to  develop  monitoring  techniques  that  capital- 
ize on  sensitive  environmental  indicators  in  zones  where 
whitebark  pine  is  a  component. 

•  More  quantifiable  parameters  and  equations  for 
building  simulation  models  for  all  projected  resource  use. 

•  More  about  habitat  requirements  in  these  forests  for 
large  mammals  such  as  bighorn  sheep,  goats,  and  moose. 

•  How  to  devise  monitoring  and  long-term  permanent 
plots  to  use  this  ecosystem  to  detect  climate  change. 

•  How  to  increase  communication  between  managers 
and  researchers. 

Landscape 

•  stand  structure  relationships  at  different  altitudes, 
latitudes,  geographic  areas,  and  ecotypes. 


310 


•  How  to  effectively  utilize  artificial  intelligence  (AI), 
geographic  information  systems  (GIS),  and  various  mod- 
els for  evaluating  landscape-scale  activities,  including 
micro-to-macro  extrapolation  methods. 

•  How  to  relate  natural  mosaics  to  human  activity 
mosaics. 

•  How  to  better  manage  information  to  more  effec- 
tively expedite  transfer  from  research  to  managers. 

Regional 

•  Information  from  a  survey  of  whitebark  pine  condi- 
tions throughout  its  range,  including  insect  and  disease 
conditions. 

•  How  to  develop  more  replicable  information  through- 
out the  whitebark  pine  range. 

•  The  history  of  whitebark  pine  throughout  its  range. 

•  Fire  history  throughout  the  whitebark  pine  range. 

Continental 

•  How  to  better  distribute  research  throughout  the 
geographic  range  of  whitebark  pine. 

Global 

•  Better  methods  of  detecting  potential  climate  change 
effects  on  whitebark  pine  survival. 

DISCUSSION 

Authors  at  the  1989  whitebark  pine  symposiimi  de- 
scribed research  needs  most  often  at  the  stand  level,  but 
needs  at  the  tree  and  forest  level  were  a  close  second  and 


third  (fig.  1).  These  are  based  on  the  frequency  of  men- 
tion at  the  various  levels.  Each  mention  reflects  an  indi- 
vidual author's  intimate  knowledge  of  a  particular  level 
and  recognition  of  research  needs  at  that  level.  A  bias 
may  exist  because  tree,  stand,  and  forest  are  the  levels 
at  which  most  forestry  research  has  traditionally  been 
conducted  (figs.  2,  3,  4).  But,  in  spite  of  emphases  at  the 
stand  and  tree  levels,  there  was  a  wide  distribution  of 
research  needs  suggested  across  the  spectnmi  from  gene 
to  global  levels  (fig.  1). 


Figure  2 — Individual  tree  growth  is  relatively  slow 
on  high-elevation  sites  where  whitebark  pine 
grows.  How  are  tree  growth  dynamics  related 
to  soil,  site,  and  insect  and  disease  conditions? 


311 


Figure  3 — Whitebark  pine  stands,  composed 
of  mature  trees  such  as  these  in  the  Absarokee 
Mountains  of  Montana,  produce  the  seeds  so 
important  to  the  survival  of  the  grizzly  bear, 
squirrel,  and  nutcracker.  What  stand  conditions 
optimize  conditions  for  these  wildlife  species? 


A  subsequent  workshop  (McCaughey  and  McDonald 
1993)  that  updated  some  of  the  information  presented 
at  the  whitebark  pine  symposium  of  1989  also  included 
a  query  of  research  needs.  The  scale  of  needs  from  gene 
to  global  level  was  essentially  the  same  as  that  found  at 
the  1989  symposium.  Research  needs  expressed  at  the 
1993  workshop  reflected  particular  concern  with  the 
threat  of  significant  blister  rust  invasion  and  mortality 
in  the  Greater  Yellowstone  Ecosystem. 


Some  of  the  authors  at  the  1989  S5rmposium  were  from 
the  management  community,  and  they  addressed  the 
knowledge  gaps  they  are  facing  when  they  have  to  make 
difficult  forest  management  decisions.  Most  managers 
felt  that  researchers  should  increase  the  scope  of  their 
research.  Although  they  saw  many  of  the  same  needs  ex- 
pressed by  the  researchers,  managers  tended  to  express 
more  concerns  about  needs  at  the  ecosystem  and  land- 
scape levels  (fig.  5).  This  is  not  surprising,  given  the  re- 
cent emphases  on  Ecosystem  Management  that  are  being 
adopted  by  most  of  the  major  public  land  management 
agencies  in  the  United  States.  This  holistic  approach  to 
management  presents  a  significant  new  challenge  to  the 
research,  education,  and  maneigement  comm\mities. 

It  is  obvious  that  there  are  enough  research  needs  in 
whitebark  pine  ecosystems  to  charter  a  major  ecological 
research  progrsun.  Fortunately,  the  evolution  of  electronic 
methods  for  handling  large  amounts  of  data,  and  develop- 
ing conceptual,  mathematical,  and  visual  models,  may 
help  speed  solutions  to  ecological  problems.  Although 
the  challenge  is  great,  the  mood  and  tools  for  tackling 
Ecosystem  Management  are  being  developed.  The  white- 
bark pine  ecosystem  is  a  good  candidate  for  testing  these 
methods. 

It  should  be  emphasized  that  the  research  needs  de- 
scribed in  this  paper  help  set  the  direction  and  priority 
of  research,  but  other  factors  come  into  play.  For  ex- 
ample, just  because  a  research  need  is  mentioned  most 
frequently  does  not  necessarily  mean  that  need  has  the 
highest  priority.  An  example  is  the  severe  mortality 
problems  associated  with  blister  rust.  If  we  cannot  solve 
the  blister  rust  mortality  problem,  some  of  the  other 
whitebark  pine  research  becomes  almost  academic.  Ide- 
ally, everyone  working  with  whitebark  pine  would  like 
to  see  a  comprehensive  and  integrated  research  program 
throughout  the  range  of  whitebark  pine  in  North  America. 


Figure  4 — Whitebark  pine  forests 
occupy  scenic  high-elevation  areas. 
Disturbances,  such  as  this  windthrow, 
create  small  openings  where  nut- 
crackers cache  whitebark  pine  seeds, 
resulting  in  subsequent  regeneration. 
What  are  the  regeneration  mecha- 
nisms and  requirements  for  whitebark 
pine? 


312 


Figure  5 — Whitebark  pine  research 
needs  from  micro  to  macro  levels  (gene 
to  global),  based  on  frequency  of  men- 
tion by  authors  from  the  research  and 


Tree  Forest        Landscape     Continental  management  communities  at  the  white- 

Gene  Stand         Ecosystem       Regional         Global  bark  pine  symposium. 


Realistically,  we  know  that  will  not  happen.  As  Lanner 
(1993)  stated  in  his  editorial  concerning  the  threat  to  the 
svirvival  of  whitebark  pine:  "Each  organization  has  a  very 
few  dedicated  researchers  putting  mere  fractions  of  their 
time  into  relevant  whitebark  pine  research,  but  the  need 
greatly  outstrips  their  ability  to  fill  it."  Lanner  recom- 
mended that  the  Forest  Service  and  Park  Service  assemble 
an  "interagency  commission  to  come  to  grips  with  the  im- 
pending disaster  now  facing  whitebark  pine,  and  to  give 
high  priority  to  that  research  and  program  implementa- 
tion deemed  most  promising."  We  have  a  national  and 
international  responsibility  to  meet  these  challenges. 

REFERENCES 

Lanner,  Ronald  M.  1993.  Is  it  doomsday  for  whitebark 
pine?  Western  Journal  of  Applied  Forestry.  8(2):  47-48. 

McCaughey,  Ward  W.;  McDonald,  Kathy  J.  1993.  Manage- 
ment of  whitebark  pine  ecosystems — an  international 


and  regional  perspective:  workshop  proceedings;  1993 
April  23;  Bozeman,  MT.  Bozeman,  MT:  Intermoun- 
tain  Research  Station,  Forestry  Sciences  Laboratory, 
Montana  State  University.  56  p. 

McCaughey,  Ward  W.;  Weaver,  T.  1990.  Reference  guide 
to  whitebark  pine.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  white- 
bark pine  ecosystems:  ecology  and  management  of  a 
high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agricvdtxire,  Forest  Service,  Intermoimtain  Re- 
search Station:  376-386. 

Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps.  1990. 
Proceedings — symposivun  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermoimtain  Research  Station. 
386  p. 


313 


PROBLEMS  OF  COMPREHENSIVE 
INVESTIGATION,  UTILIZATION,  AND 
REPRODUCTION  OF  RUSSIAN  CEDAR 
PINE  FORESTS 

Vladislav  N.  Vorobjev 
Nina  A.  Vorobjeva 


Abstract— Cedar  pine  forests  occupy  vast  areas  in  Russia  and  eire 
composed  of  three  types  of  pines:  the  Siberian  stone  (or  cedar) 
pine  (Pinus  sibirica  Du  Tour),  Korean  pine  (Pinus  koraiensis  Sieb. 
et  Zucc),  and  mountain  pine  {Finns  pumila  [Pall.]  Kegel).  All  of 
these  are  stone  pines  and  are  pioneer  tree  species.  Most  of  the  pa- 
per discusses  Siberian  stone  pine,  the  main  stone  pine  species  in 
Russia.  These  forests  not  only  provide  valuable  raw  materials,  but 
other  ecological  values  are  increasingly  being  recognized.  This  pa- 
per describes  cone  and  seed  development  and  production,  artificial 
reproduction,  site  and  stand  characteristics,  seed  morphology  and 
chemistry,  climatology,  tree  growth,  and  other  characteristics  of 
these  stone  pine  trees  and  forests.  Management  implications  are 
also  discussed. 


Cedar  pine  is  the  commonly  used  term  for  stone  pine  for- 
ests found  in  Russia.  These  forests  are  composed  of  three 
types  of  cedar  pines:  Siberian  stone  (or  cedar)  pine  {Pinus 
sibirica  Du  Tour),  Korean  pine  {Pinus  koraiensis  Sieb.  et 
Zucc),  and  mountain  pine  {Pinus  pumila  [Pall.]  Kegel).  All 
these  occupy  about  40  million  ha,  with  Siberian  stone  pine 
accounting  for  90  percent  of  the  total. 

Cedar  pine  forests  are  of  paramount  importance  for 
Russia.  Cedars  not  only  provide  valuable  raw  materials, 
but  they  also  play  an  ever-increasing  role  in  ecology.  Their 
stands  are  critical  for  conservation  of  water  resources  (on 
mountains  and  near  sources  of  great  Siberian  rivers  the 
Ob,  the  Irtish,  the  Yenisei,  and  also  around  Baikal  Lake) 
and  they  also  serve  to  stabilize  swamping  processes  (in  the 
West  Siberian  Plain).  Cedar  pines  are  pioneer  tree  species 
known  for  their  soil-cover  protection  function,  especially 
for  the  subalpine  zone. 

ECOLOGICAL  FORMS 

Ecological  conditions  for  Siberi£in  stone  pine  growth  vary 
by  latitude  and  altitude.  The  trees  of  typical  form  grow  at 
the  altitude  of  the  subalpine  zone  in  North  Altai  where  the 
elevation  is  1,500-1,600  m.  Other  ecological  forms  include 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11, 1992. 

Vladislav  N.  Vorobjev  is  Head  and  Nina  A.  Vorobjeva  is  a  Researcher  of 
the  Institute  for  Ecology  of  Natural  Complexes,  The  Academy  of  Sciences, 
Siberian  Division,  Academichesky  pr.  2,  Academgorodok,  Tomsk,  634055, 
Russia. 


the  most  important  one,  a  moimtain  form  (f.  coranans).  It 
is  unUke  trees  of  typical  form  and  is  comprised  of  second- 
value  trees  with  dense  crowns,  especially  in  the  zone  of  the 
male  cone  growth.  Branches  of  this  type  of  crown  usually 
begin  at  ground  level.  The  female  zone  of  the  crowns  com- 
prises less  than  30  percent  of  total  crown  length.  Develop- 
ment of  generative  £ind  vegetative  organs  of  this  mountain 
form  of  Siberian  stone  pine  begins  a  month  later  than  those 
trees  in  the  low  part  of  the  mountains.  Cone  crops  are  rare 
and  small. 

The  mean  length  of  f.  cororuins  cones  is  62-64  mm  in  years 
with  good  crops  £ind  43-45  mm  in  years  with  bad  crops.  In 
bad  crop  years  the  cones  are  small  and  ball-like  with  a  mean 
width  of  38-40  mm  and  usually  20-30  large  seeds  (the  weight 
of  1,000  seeds  is  260-280  g).  The  cones  have  scales  with  a 
red  hue.  Most  of  the  scales  (40  percent)  have  a  flat  apophy- 
sis. The  weight  of  1,000  seeds  is  170  ±  30  g  in  f.  cororuins 
and  240  ±  70  g  in  typical  form  trees  in  years  with  good  crops. 
Seeding  quality  of  f.  coronans  seeds  is  low,  seed  germinat- 
ing capacity  is  less  than  30-40  percent,  and  one-third  of  the 
seeds  have  no  embryo. 

F.  coronans  reproduction  very  seldom  results  from  its 
own  seeds.  Instead,  nutcrackers  {Nucifraga  caryocatactes 
macrorhynchos  Brehm)  carry  seeds  from  lower  elevation 
stands  of  the  t5T)ical  form  of  Siberian  stone  pine  to  the  up- 
per part  of  the  mountains.  The  nutcracker  seed  distribution 
plays  a  significant  role  in  the  preservation  off.  coronans 
areas.  Correlation  of  this  way  of  reproduction  and  growth 
of  f.  coronans  with  other  subalpine  forms  has  not  been  ex- 
actly determined.  This  problem  is  especially  important  for 
investigating  ecology  of  growth  and  generative  development 
of  Siberian  stone  pine  in  the  different  altitude  zones. 

Another  ecological  form  of  Siberian  stone  pine  is  f.  nana 
(elevation  is  1,900-2,000  m).  Trees  here  grow  to  about  10  m 
height,  disperse  as  single  trees  or  cliunps,  and  do  not  form 
stands.  The  boundary  of  tree  life  form  £ind  the  generative 
boundary  of  the  species  is  here.  Trees  here  bear  few  Siberian 
stone  pine  cones  and  have  only  about  15-20  female  shoots 
in  the  crown.  Oval  cones  are  about  50  mm  in  length,  have 
no  more  than  20-30  seeds,  and  they  mature  only  rarely. 

The  upper  end  of  the  ecological  row  is  f.  humistrata  (other 
name  is  f.  depressa).  It  is  sterile.  This  form  is  a  small  tree 
or  shrub  2-3  m  in  height.  F.  humistrata  forms  the  upper 
forest  boundary  (elevation  is  2,000-2,400  m). 

Mean  height  of  trees  of  these  four  ecological  forms  of 
Siberian  stone  pine  decreases  with  increasing  elevation:  35, 
17,  10,  and  2  m.  Characteristics  of  the  typical  form  and  f. 
coronans  stands  at  different  elevations  are  shown  in  table  1. 


314 


Table  1 — Characteristics  of  stands  of  the  Kiga  ecological  line 


Number 
of 


sample 

Site  index, 

Tree 

Stand 

Stand 

Mean 

Mean 

Growing 

nint 

TOicsi  lype 

CIcVallOn 

story 

uonieni 

□ensiiy 

age 

neigni 

aiameier 

stock 

m 

Year 

m 

cm 

nf/ha 

2a 

1,  large-grassy 

450 

1 

3PS(I) 

0.15 

300 

35 

80 

90 

7PS(II) 

.44 

210 

32 

58 

296 

2 

9AS 

.36 

130 

17 

25 

90 

IBs 

.03 

90 

20 

39 

10 

Unit  PS 

90 

12 

10 

2 

8a 

II,  green-mossy  fern 

1,250 

1 

10PS 

.76 

270 

29 

67 

470 

2 

10AS 

.35 

130 

19 

21 

103 

Unit  PS 

130 

13 

13 

4 

13 

V,  large-grassy 

1,800 

1 

10PS 

1.15 

210 

17 

37 

340 

with  Leuzea 
carthamoides 


'Abbreviations:  PS — Pinus  sibirica  Du  Tour;  AS — Abies  sibirica  Ldb.;  Bs— Sefu/a  sp. 


The  change  of  stand  characteristics  and  the  formation  of 
Siberian  stone  pine  ecological  forms  follows  climatic  indexes. 
Yearly  air  temperature  changes  from  +1.4  °C  to  -2.8  °C,  the 
sum  of  temperatiires  above  10  °C  decreases  from  1,400  °C 
to  630  °C,  and  the  sum  of  precipitation  increases  from  450 
to  980  mm  with  elevation  increases  from  450  to  2,000  m. 

The  isolation  of  Siberian  stone  pine  ecological  form  is  con- 
firmed by  hemotaxonomy  data  (Vorobjev  and  others  1971). 
The  investigation  of  monoterpene  contents  in  oleoresin 
shows  differences  of  quantity  between  the  typical  form  and 
the  f.  nana  trees  as  well  as  in  the  different  altitude  zones 
(table  2). 

a-pinene  and  (3-pinene  are  decreased  and  the  quantity 
of  A^-karene  and  P-phellandrene  is  increased  in  the  typical 
form  of  Siberian  stone  pine  with  increasing  elevation  in  the 
moimtains.  We  observed  the  analogical  change  of  monoter- 
pene content  of  Siberian  stone  pine  trees  in  connection  with 
resin-tapping  done  for  the  purpose  of  obtaining  resin  sub- 
stances for  different  industrial  tasks.  This  is  related  usu- 
ally to  increasing  tree  stability  due  to  influence  of  different 
factors,  including  the  enhancing  of  oleoresin-forming  activity. 


It  is  believed  that  distinguishing  monoterpene  content  in 
f.  nana  trees,  unlike  typical  tree  form,  will  be  analogical. 
That  is,  the  quantity  of  a-pinene  and  P-pinene  will  be  less 
and  A^-karene  will  be  more.  It  appears  that  the  monoter- 
pene content  in  the  oleoresin  changes  in  the  other  direction. 

In  f.  nana  the  contents  of  a-pinene,  and  P-pinene  espe- 
cially, are  more  and  A^-karene  and  P-phellandrene  are  less 
when  compared  with  data  for  the  typical  form  (elevation  is 
1,250  m). 

These  changes  have  formed  the  simplified  monoterpene 
content  in  f.  nana  oleoresin.  Three  basic  hydrocarbons 
(a-pinene,  P-pinene,  and  A^-karene)  are  contained  in  f.  nana. 
The  total  sum  is  97  percent  in  f.  nana  and  88.6  percent  in 
the  typical  form  of  Siberian  stone  pine.  The  other  direction 
of  synthesis  of  monoterpene  and  its  simplified  content  in 
f.  nana  is  possibly  related  to  the  change  for  the  worse  of  eco- 
logical condition. 

Our  study  provides  perspective  in  this  direction.  Other 
studies  are  needed  to  develop  this  perspective  in  different 
regions  and  in  different  species  of  cedar  pines.  These  types 
of  investigations  are  appropriate,  especially  in  mountains 
where  ecological  forms  can  be  distinguished  very  easily. 


Table  2 — Monoterpene  content  in  oleoresin  of  Pinus  sibirica  Du 

Tour  Typical  Form  (T)  and  Pinus  sibirica  Du  Tour  f.  nana 
Beissn.  (N)  (percent) 


T   N 


450  m 

1,250  m 

(1,900  m 

Monoterpene 

elevation 

elevation 

elevation) 

a-pinene 

50.5  ±  1.5 

39.1  ±  1.2 

47.9  ±  1 .5 

Kamphene 

.8  ±  .2 

1.2  ±  .2 

.7  ±  .2 

p-pinene 

11.0  ±  .4 

9.6  ±  .4 

^20.3  ±  .6 

A^-karene 

27.6  +  .8 

39.5  ±  1.2 

'28.8  +  .8 

Mirtsen 

1 .5  +  .2 

1.4  ±  .2 

'.6  ±  .2 

Limonene 

2.0  ±  .2 

.6  ±  .2 

'.5  ±  .2 

P-phellandrene 

6.4  ±  .3 

8.4  ±  .4 

\2  ±  .1 

'Values  are  very  different  from  mean  in  typical  form. 


ECOLOGICAL  FACTOR  INFLUENCE 

The  influence  of  ecological  factors  on  development  of 
Siberian  stone  pine  trees  and  their  physiological  processes 
is  illustrated  by  a  good  visible  example  of  lipid  accumulation 
in  seeds.  Table  3  shows  accimaulative  dynamics  in  seeds  for 
trees  of  typical  form,  f.  coronans,  and  f.  nana  (elevations 
are  450,  1,650,  and  1,850  m,  respectively). 

It  is  obvious  that  lipid  acciunulation  is  significantly  late 
in  the  subalpine  form.  The  seeds  have  standard  lipid  con- 
tent (60  percent)  at  the  end  of  July  (7/29)  in  the  typical 
form  of  Siberian  stone  pine,  in  the  third  week  of  September 
(9/22)  in  f  coronans,  and  at  the  beginning  of  October  (10/2) 
in  f  nana.  The  lipid  content  of  the  subalpine  ecological  form 
reaches  the  optimal  value  in  crop  years.  It  confirms  the 
physiological  capacity  of  its  seed  for  reproduction. 


315 


Table  3 — Dynamics  of  lipid  content  in  Siberian  stone  pine  seeds  in  mature  process  (percent  of  absolute  dry  substance) 


Elevation      20.6        26.6  2.7         9.7        22.7        29.7  6.8        13.8        13.9        22.9        27.9        2.10  10.1 


m 

450  8.1  21.6  29.2  47.6  51.2  61.6  62.2  62.5  64.2  63.5  64.1           —  — 

1,650  —  3.0  4.9  7.6  17.4  29.6  43.3  57.0  60.3  61.0  61.6  64.6  62.5 

1,850  —  2.5  4.9  6.9  17.4  27.0  42.4  52.2  53.2  56.4  58.4  61.1  57.5 


On  the  whole,  the  dynamics  of  the  maturing  of  Siberian 
stone  pine  seeds  are  characterized  by  slow  lipid  accumula- 
tion, then  the  intensification  of  this  process  in  July,  and 
some  decreasing  of  total  lipid  content  after  the  determina- 
tion of  optimal  correlation  of  fatty  acids.  This  phenomenon, 
known  as  "over-ripening,"  is  characteristic  of  Siberian  stone 
pine  as  well  as  many  oily  plants. 

The  Siberian  stone  pine  and  other  cedar  pine  seeds  have 
great  value  as  food  in  Russia.  General  physical  and  chemi- 
cal characteristics  of  Siberian  stone  pine  seed  have  been 
published  in  a  collective  monography  (Vorobjev  and  others 
1979)  and  also  in  some  papers  (Rush  1974;  Rush  and 
Lizunova  1969).  This  is  shown  in  table  4. 

Together  with  the  high  lipid  content  the  seeds  have  many 
amino  acids,  including  irreplaceable  acids  as  well  as  many 
macro-  and  micronutrient  elements,  and  especially  valuable 
vitamins  B^,  E,  and  F.  The  Siberian  stone  pine  seeds  for 
B-vitamin  activity  are  more  preferable  then  the  other  cedar 
pines  and  the  oily  plants  (table  5).  Also,  Siberian  stone  pine 
seeds  are  valuable  for  macro-  and  micronutrient  elements 
(table  6). 

The  quantitative  contents  of  known  elements  of  Siberian 
stone  pine  seeds  are  similar  to  mountain  pine  seeds,  and 
they  have  more  phosphorus,  iodine,  and  cobalt  than  Korean 
pine  seeds.  The  significant  quantity  of  potassium,  sodixmi, 
copper,  and  zinc  in  the  Siberian  stone  pine  seeds  determines 
its  high  quality. 

The  high  content  of  iodine  in  Siberian  stone  pine  and 
mountain  pine  seeds  has  special  importance  for  those  re- 
gions where  trees  have  endemic  goiter. 

A  comparison  of  the  seed  content  fi'om  the  plain  and  moun- 
tain Siberian  stone  pine  forests  shows  an  increased  content 
of  manganese,  silicon,  boron,  nickel,  and  phosphorus,  as  well 
as  iron  in  the  middle  part  of  the  moimtains.  The  seeds  fi'om 
different  plant  regions  can  be  distinguished  by  increased  con- 
tent of  potassium,  zinc,  molybdeniun,  alimiiniimi,  and  iodine. 
In  the  upper  part  of  the  mountains  the  content  of  phospho- 
rus, copper,  zinc,  and  boron,  the  most  important  elements 
determining  viability  of  seeds,  is  decreased.  The  content  of 
magnesium,  iron,  and  silicon  also  is  less  diminished  high 
in  the  mountains. 

On  the  whole,  the  changes  of  mineral  content  of  seeds 
are  related  to  their  physiological  and  biochemical  matura- 
tion, ecological  conditions  of  tree  growth,  and  geochemical 
indexes  of  cedar  pine  in  the  widespread  regions. 

CONE  AND  SEED  CROPS 

The  investigations  of  cone  and  seed  crop  dynamics  have 
a  special  meaning  for  study  of  growth  and  generative  devel- 
opment of  Siberian  stone  pine  trees.  The  method  of  retro- 
spective calculation  of  crops  for  the  last  100  years  or  more 


Table  4 — Characteristics  of  Siberian  stone  pine  seed 


Characteristic  Mean  Range 


Physical  indexes 


Length  (mm) 

10.4000 

8.9000 

-  11.9000 

Width  (mm) 

7.7000 

6.0000 

-  8.7000 

Seed  weight  (mg) 

200.0000 

153.0000 

-317.0000 

Nucleus  weight  (mg) 

96.0000 

78.0000 

-125.0000 

Shell  (percent  of  weight) 

51 .6000 

51 .5000 

-  58.9000 

Nucleus  (percent  of  weight) 

46.6000 

41 .0000 

-  48.6000 

Seed  coat  (percent  of  weight) 

1.8000 

1.5000 

-  2.4000 

Density  (g/cm') 

.8700 

.8500 

.8800 

Specific  volume  (g/cm^) 

1.1500 

1.1400 

-  1.2000 

Chemical  indexes 

Nitrogen-bearing  substance  (percent  of 

absolute  dry  substance): 

total  nitrogen 

2.9000 

2.0800 

-  3.2700 

protein  nitrogen 

2.6100 

1.9300 

-  3.0100 

nonprotein  nitrogen 

.2900 

.1500 

.3800 

Total  nitrogen  of  unlipid  remainder 

(percent  of  absolute  dry  substance) 

7.7000 

7.5800 

-  7.8800 

protein  nitrogen 

4.8300 

4.6000 

-  5.1500 

nonprotein  nitrogen 

.7000 

.6000 

.8300 

nitrogen  of  dense  remainder 

2.1700 

1.8800 

-  2.3800 

Free  amino  acids  (mg  percent  of  absolute  dry  substances) 

Cyctine 

10.5000 

Traces 

-  19.5000 

Lisine 

15.3000 

2.2000 

-  35.7000 

Histidine 

17.1000 

13.0000 

-  25.7000 

Arginine 

15.5000 

12.0000 

-  26.2000 

Aspartic  acid 

23.3000 

15.5000 

-  30.2000 

Serine 

9.8000 

3.5000 

-  14.3000 

Glycine 

4.5000 

Traces 

-  11.2000 

Glutamic  acid 

50.3000 

46.4000 

-  87.9000 

Threonin 

16.7000 

10.9000 

-  18.7000 

Alanine 

30.7000 

28.7000 

-  44.4000 

Proline 

24.1000 

17.9000 

-  35.7000 

Tyrosine 

17.3000 

14.5000 

-  32.1000 

Methionine 

13.4000 

Traces 

-  42.9000 

Valine 

8.4000 

5.8000 

-  13.0000 

Tryptophan 

Traces 

Traces 

-  Traces 

Phenylalanine 

13.5000 

Traces 

-  34.0000 

Leucine-isoleucine 

14.2000 

8.6000 

-  25.6000 

Protein  acids 

Lisine 

1.2000 

0.6000 

-     1 .4000 

Histidine 

.9000 

.6000 

-  1.4000 

Arginine 

5.9000 

3.1000 

-  7.5000 

Aspartic  acid 

4.9000 

1.8000 

-  5.9000 

Threonine 

1.0000 

.8000 

-    1 .5000 

Serine 

2.1000 

1.6000 

-  2.1000 

Glutamic  acid 

4.6000 

2.9000 

-  7.3000 

Proline 

1 .6000 

.7000 

-  2.2000 

Glycine 

1.7000 

.7000 

-  1.7000 

Alanine 

1.9000 

.7000 

-  2.1000 

Valine 

1.1000 

.9000 

-  2.2000 

Methionine 

.6000 

.2000 

.8000 

Isoleucine 

1.6000 

.8000 

-  2.0000 

Leucine 

2.0000 

1.3000 

-  3.1000 

Tyrosine 

1.5000 

.8000 

-  1.7000 

Phenylalanine 

1.4000 

.8000 

-  1.6000 

Tryptophan 

.6000 

0.6000 

(con.) 


316 


Table  4  (Con.) 


Characteristic  Mean  Range 

Contents  of  carbohydrate  nature  (percent  of  absolute  dry  substance) 


Glucans: 


cellular  tissue 

2.2000 

1.9000 

-  2.4000 

starch 

4.5000 

1.9000 

-  8.2000 

dextrine 

2.3000 

2.1000 

-  2.5000 

pentosanes 

1.8000 

1 .6000 

-  2.2000 

Sum  of  easily  soluted  carbohydrates: 

6.2000 

3  3000 

-  14  3000 

sucrose 

5.1000 

2.1000 

raffinose 

3.4000 

4  7nnn 

glucose 

.1000 

.1000 

'^000 

.\J\J\J\J 

fructose 

.2000 

.1000 

.8000 

Lipids 

Lipid  content  (percent  of  absolute 

dry  substance) 

64.0000 

50.0000 

-  76.0000 

Acid  content  in  oil  (percent) 

nonsaponificated  substances 

1.1200 

.7800 

-  1.3500 

saturated  acids 

5.7600 

5.1300 

-  6.4200 

unsaturated  acids: 

oleic  acid 

15.7600 

1 1 .3300 

-  22.2200 

llnoleic  acid 

57.2400 

53.1400 

-  59.1100 

linolenic  acid 

21.2200 

16.5800 

-  24.6400 

Mineral  substances  (P-Fe  in  mg  percent,  Mn-  in  mg/kg) 

P 

481.8000 

428.4000 

-716.1000 

Mg 

529.4000 

258.9000 

-559.8000 

K 

489.3000 

350.9000 

-503.3000 

Na 

107.1000 

83.9000 

-114.3000 

Ca 

48.4000 

35.1000 

-  49.6000 

Fe 

2.3000 

2.0000 

-  4.9000 

Mn 

5.4000 

5.2000 

-  11.3000 

Cu 

1.5000 

1.3000 

-  3.9000 

Zn 

12.1000 

7.1000 

-  15.1000 

Mo 

.2000 

.1000 

.3000 

Si 

2.3000 

2.0000 

-  3.7000 

Al 

5.3000 

3.0000 

-  6.4000 

1 

.5000 

.3000 

.9000 

B 

.0002 

.0001 

.0017 

Ni 

.0300 

.0090 

.0470 

Go 

.0500 

.0390 

.0980 

Pb 

.0300 

.0270 

.0640 

Sr 

.0004 

.0001 

.0029 

Ag 

.0300 

.0190 

.0910 

Aches  content  (percent) 

2.5000 

2.4000 

-  2.7000 

Chemical  indexes  of  oil 

Acid  number  (mg  KOH) 

1.4100 

1.1800 

-  1.7100 

Saponification  value  (mg  KOH) 

196.9000 

189.7000 

-201.3000 

Relhert-Meissl's  number 

1.3300 

1.1100 

-  1.5400 

Polanske's  number 

.4100 

.2900 

.5900 

Cube's  iodine  number 

166.6000 

158.2000 

-177.2000 

Rhodanic  number 

97.1000 

94.2000 

-100.3000 

Phosphatide  content  (percent) 

1.3400 

1.0000 

-  1.6500 

Vitamins  (mg  percent): 

A  (carotene) 

Traces 

Traces 

B,  (thiamine) 

.6500 

.5500 

.6900 

Bj  (riboflavin) 

.9300 

.8400 

-  1.2100 

C  (ascorbic  acid) 

Traces 

Traces 

E  (tocopherols): 

in  seeds 

10.1000 

9.2000 

-  32.8000 

in  oil 

54.8000 

50.1000 

-  65.0000 

F  (unsaturated  fatty  acid)  (percent) 

94.3000 

93.6000 

-  94.9000 

was  worked  out  in  our  institute  for  ecology  of  natural  com- 
plexes in  Tomsk  (Vorobjev  1979).  Data  for  Siberian  stone 
pine  crop  dynamics  for  the  latitudinal  profile  of  the  West 
Siberia  area  can  be  obtained  by  using  this  method  (fig.  1). 

It  has  been  determined  that  the  cone  crop  dynamics  have 
cycles  of  different  duration:  for  3  years,  10  or  11  years,  28 


Table  5 — Vitamin  content  in  seeds  (mg  percent  of  absolute  dry 
substance) 


Plant  spocies 

Thiamine 

Riboflavin 

Total 

Pin  us  sibirica  Du  Tour 

0.242 

0.933 

1.175 

Amygdalus  nana  L. 

.111 

.730 

.841 

Pistacia  vera  L. 

.149 

.639 

.788 

Pinus  koraiensis  Sieb.  et  Zuss. 

.236 

.210 

.446 

Pinuspumila  (Pall.)  Regel 

.049 

.263 

.312 

Juglans  regia  L. 

.096 

.186 

.282 

or  33  years,  and  longer.  Relationship  of  the  cycles  is  very 
complicated  in  different  latitudes  and  within  each  latitude. 
Examples  of  likeness  and  unlikeness  of  longer  cycles  are 
available.  The  differences  are  observed  especially  in  shorter 
cycles.  Except  for  the  direct  effect  of  weather  condition  on 
the  development  of  generative  organs,  solar  activity  influ- 
ences crop  dynamics.  This  problem  is  studied  widely  in  den- 
drochronology. Studies  of  the  effect  of  solar  activity  on  seed- 
bearing  of  forest  trees,  including  Siberian  stone  pine,  is  at 
the  initial  stage. 

Study  of  the  crop  dynamics  is  planned  for  long-range 
predictions.  You  can  see  in  figure  2  that  solar  activity  has 
a  direct  and  positive  influence  on  shoot  growth.  After  its 
active  increment  the  cycle  of  good  crops  of  Siberian  stone 
pine  begins. 

MORPHOPHYSIOLOGY 

Knowledge  of  the  morphophysiological  state  of  separate 
tissues  and  organs  during  the  processes  of  morphogenesis 
is  one  of  the  basic  conditions  needed  to  determine  the  role 
of  growth  in  the  cone-bearing  and  regulation  of  periodicity 
of  Siberian  stone  pine  crops.  It  is  necessary  to  study  the 
morphological  and  physiological  regularities  of  growth  and 


Table  6 — Content  of  macronutrient  and  micronutrient  elements  In 
cedar  pines  (Russ  1974) 

Element^     Pinus  sibirica     Pinuspumila     Pinus  koraiensis 

P 

Mg 
K 

Na 
Ca 
Fe 
Mn 
Cu 
Zn 
Mo 
Si 
Al 
I 

B 
Ni 
Co 
Pb 
Sr 

Ag  

'P-Fe  in  mg  percent,  Mn-Ag  in  mg/kg. 


481.8000 

586.2000 

341 .9000 

529.7000 

317.4000 

682.4000 

489.3000 

452.8000 

549.6000 

107.1000 

113.8000 

278.4000 

48.4000 

49.6000 

82.4000 

2.3000 

4.1000 

11.3000 

5.4260 

8.3120 

9.0110 

1 .4790 

1.8150 

14.2130 

12.1320 

8.6520 

11.7160 

.1730 

.0930 

2.3230 

3.1440 

8.6520 

5.2780 

8.1560 

9.9130 

.4590 

.6820 

.0170 

.0002 

.0001 

.0470 

.0110 

.1610 

.0470 

.1430 

.0380 

.0290 

.0031 

.0004 

.0080 

.0001 

.0290 

.0210 

317 


Year 


Figure  1 — Cyclicity  of  Siberian  stone  pine  crops 
for  latitude  lines  of  West  Siberia:  A— 67  NL,  B— 60, 
C— 56  NL. 


sexualization  and  their  relationship  as  a  condition  that 
switches  one  development  program  to  another.  Modern 
views  consider  this  switching  the  result  of  trophic,  hor- 
monal, and  genetic  interactions  (Chaylakhyan  1984). 

Participation  of  each  of  these  factors  in  determining 
Siberian  stone  pine  sexualization  was  recently  proved  by 
the  peculiar  physiological  and  biochemical  characteristics 
of  the  different  sex  shoots.  These  data  show  that  the  pe- 
culiarities of  generative  shoots  are  a  function  of  the  differ- 
ences between  hormonal  and  trophic  substances  in  total 
content,  correlation  between  their  separate  form,  compo- 
sition of  quality,  and  the  seasonal  dynamics  rhythm. 

The  same  changes  are  also  found  in  cases  of  increasing 
or  decreasing  of  sexual  display  in  shoots  or  the  total  organ- 
ism. For  example,  increasing  female  shoot  reproduction  in 
our  study  was  correlated  with  decreases  of  oligosaccharides 
in  the  needles  and  increasing  content  in  the  shoot  axis  in 


1890   1900   1910  1920  1930  1940    1950  1960 

Year 


Figure  2— Solar  activity  (1),  apical  growth  of  fe- 
male shoots  (2),  and  crop  value  of  Siberian  stone 
pine  (3)  in  the  subalpine  subzone  in  Gorny  Altai. 


the  period  of  vegetation  end  (Vorobjeva  1973).  The  degree 
of  the  sexual  signs  and,  therefore,  the  cone-bearing  state  of 
Scotch  pine,  are  correlated  with  dynamics,  content,  metabo- 
lism direction,  and  intensity  of  differences  in  needles,  buds, 
and  shoot  axes.  But  these  alterations  are  not  clearly  related 
to  quantitative  changes  of  physiological  characteristics 
(Mosin  and  Savina  1985;  Samsonova  and  Bolgova  1985). 

This  is  all  shown  by  small  studies  of  the  individual  role 
of  each  substance  and  the  order  of  their  action  in  the  pro- 
cesses of  sexualization.  There  is  a  need  to  mark  the  ab- 
sence of  correlation  between  biochemical  characteristics 
and  the  morphological  state  of  the  organs  and  organism. 
For  example,  the  carbohydrate  accumulation  in  buds  for 
the  initiation  of  generative  primordia  or  in  shoot  axes  in 
the  finished  growth  period  may  be  a  result  of  change  in  the 
rate  of  the  following  stages  of  growth.  Both  of  these  pro- 
cesses are  factors  that  lead  to  sexualization.  But  the  asso- 
ciation of  changes  in  this  factor  with  changes  of  the  bio- 
chemical conditions  do  not  give  concrete  results.  Moreover, 
evaluation  of  changes  of  biochemical  conditions  as  only  one 
aspect  of  the  metabolism  prevents  display  of  the  processes 
leading  directly  to  sexualization.  This  situation  does  not 
allow  use  of  physiological  characteristics  as  the  mark.  There- 
fore, it  is  necessary  to  intensify  work  in  this  direction  to  de- 
termine any  connection  with  morphology. 

Besides,  the  characteristics  of  generative  shoots  that  we 
have  today  reflect  the  svunmary  metabolism  changes  imder 
influence  of  initiation  of  the  floral  stems,  development  of 
generative  structure,  and  the  generative  load.  This  also 
does  not  permit  display  of  the  conditions  of  each  stage  of 
morphogenesis. 

The  carbohydrate  metabolism  and  morphological  stage  of 
apical  female  shoots  were  studied  in  common.  This  showed 
that  the  nutrition  of  the  1-  and  2-year  cones  change  the 
content,  the  activity,  and  the  tiun  of  direction  of  the  soluble 
carbohydrate  reserve  during  the  processes  of  growth  or  res- 
ervation. Such  changes  are  different  during  the  develop- 
ment of  the  reproductive  structures  and  growth  of  shoots. 
These  changes  do  not  influence  the  direction  of  morphogen- 
esis but  make  the  stage  of  development  longer.  The  mean- 
ing of  this,  however,  is  not  clear  because  the  boundary  value 
of  growth  for  the  display  of  the  sex  is  not  known.  This  knowl- 
edge is  needed  for  each  form  of  trees  and  stage  of  ontogen- 
esis. Knowledge  of  the  ontogenesis  behavior  of  each  pheno- 
type  is  necessary  to  objectively  mark  their  productivity  and 
posterity  growth. 

Knowledge  about  morphological  changes  of  individual 
forms  of  trees  is  also  limited.  Analyses  of  the  relationship 
of  growth  and  sex  of  the  shoot  are  confirmed  by  the  correla- 
tion between  sexualization  with  all  values  composing  the 
growth,  and  the  absence  of  direct  dependence  of  the  differ- 
ences in  the  rate  of  variability  on  the  stage  as  well  as  the 
critical  value  for  change  of  stage  (table  7).  The  role  of  each 
of  the  factors  is  inconsistent.  Sex  type  of  the  tree  has  no 
direct  correlation  vnth  the  growth  characteristics.  But  in- 
direct correlation  is  the  result  of  interaction  of  the  growth 
with  other  organized  factors  (Fx). 

From  this  point  of  view,  the  data  analysis  says  much  about 
metamere  variability  of  vegetative  development  and  shorter 
morphogenesis  of  female  type  trees.  These  trees  also  con- 
served a  higher  rate  of  evocation  changes  in  the  beginning 
of  reproduction. 


318 


Table  7 — Growth  of  shoots  on  trees  of  different  sexual  types 


Tree  Fa 


Shoot  Fb 

Female       Male  Asexual 

Ffact 

Branch  age  (year) 

Vegetative 

41 .20          37.80  54.40 

_ 

Vegetative 

69.30          74.40  88.90 

— 

Female 

80.80          86.10  100.60 

Axis  length  (cm) 

Vegetative 

3.19           2.53  3.14 

Fa 

-  2.18 

Vegetative 

3.85           8.26  8.65 

Fb 

-53.80 

Female 

11.16          13.54  14.94 

rx 

-  y.ao 

Axis  diameter  (cm) 

Vegetative 

0.53           0.50  0.53 

Fa 

-  0.47 

Vegetative 

.70             .79  .84 

Fb 

-59.10 

Female 

1 .41            1 .46           1 .27 

rX 

1                      fill            ■  11 

Parenchyma  of  bark/medulla  (B/M) 

Vegetative 

5.50           6.50  7.10 

Vegetative 

4.10           3.80  3.80 

— 

Female 

1.80           2.10  1.90 

Brachybiasts  on  axis 

Vegetative 

18.70          18.80  19.30 

Fa 

-  2.04 

Vegetative 

21 .00          37.80  45.90 

Fb 

-27.50 

Female 

50.80          63.70  71.80 

Fx 

-  6.25 

Auxibiasts  on  axis  of  first  bud 

Vegetative 

0.50           0.80  0.56 

Fa 

-   1 .46 

Vegetative 

.75           2.10  1.80 

Fb 

-27.50 

Female 

.80           1 .90           1 .40 

Fx 

-  6.25 

Auxiblaston  on  axis  of  second  bud 

Female 

2.40           2.60  2.50 

Fa 

-  2.04 

Generative  germs  on  axis 

Female 

2.70           2.10  2.00 

Higher  metamere  variability  correlates  with  increased 
frequency  of  double  bud  formation  during  the  vegetation 
season  in  the  period  of  virgin  development.  The  biological 
regularities  of  the  relationship  of  resin-forming,  and  genera- 
tive and  growth  processes  in  Siberian  stone  pine  are  used 
as  a  theoretical  principle  for  working  out  optimal  combina- 
tions of  elements  for  industrial  complexes. 

FOREST  MANAGEMENT 

The  comprehensive  investigation  of  Siberian  stone  pine 
forests  of  many  regions  of  Siberia  by  its  scientists,  includ- 
ing the  authors  of  this  paper,  for  the  last  30  years  shows 
that  management  in  these  forests  must  be  distinguished 
from  management  of  other  forests. 

Traditional  practices  of  forest  management  are  planting, 
receiving,  utilization,  and  reproduction  of  wood.  Now  added 
to  these  purposes  are  the  observations  of  many  ecological  de- 
mands that  do  not  change  the  classic  scheme  and  techniques 
of  forest  management.  Previously,  the  main  aim  of  this  for- 
est management  was  wood  production.  Its  basic  elements 
were  mainly  technical  exploitability  ages  and  felling  ages. 


traditional  cleaning  cutting  and  major  produce,  clear  cut- 
ting, and  "continuous"  artificial  regeneration. 

Mostly,  these  elements  are  unfit  for  Siberian  stone  pine 
forests.  New  theoretical  principles  and  new  practices  of 
forest  management  are  required. 

First  of  all,  the  purpose  of  forest  management  in  Siberian 
stone  pine  forests  is  not  the  wood  only,  but  the  comprehen- 
sive utilization  of  all  forest  resources  including  wood,  nuts, 
oleoresin,  berries,  mushrooms,  furs,  game,  and  ecological 
and  esthetic  values.  Technical  exploitability  ages,  felling 
ages,  and  major  produce  should  not  be  used  as  criteria  for 
marking.  Every  tree,  every  stand  must  be  comprehensively 
examined,  but  not  only  on  the  basis  of  wood  growing  stock. 
Trees  must  be  felled  in  conformance  with  these  value  esti- 
mations. Thus,  qualitative  evaluations  of  every  stand  and 
every  tree  in  Siberian  stone  pine  forests  must  be  made,  but 
the  middle-statistical  method  of  forest  felling  on  the  basis 
of  the  correlation  of  age,  growing-stock,  and  increment 
wood  is  not  used. 

The  classic  methods  of  forest  management  in  Siberian 
stone  pine  forests  are  not  acceptable  because  Siberian  stone 
pine  is  not  an  ordinary  tree.  This  tree  has  valuable  nuts 
and  many  other  specific  properties.  Based  on  felling  ages 
of  Siberian  stone  pine  under  previous  management  prin- 
ciples, this  tree  must  be  felled  at  160  years  old  to  obtain 
the  best  wood.  But  Siberian  stone  pine's  most  intensive 
cone-bear  occurs  in  trees  about  160  years  old.  Because  of 
this  connection  it  is  not  possible  to  fell  the  Siberian  stone 
pine  at  this  age.  Thus,  felling  schedules  must  be  deter- 
mined by  the  state  of  cone  bearing  and  other  comprehen- 
sive properties. 

The  new  theoretical  principles  of  forest  management  in 
Siberian  stone  pine  forests  have  been  developed  by  our  Insti- 
tute. They  are  published  in  some  monographs  (Isaev  1985; 
Vorobjev  1983)  and  as  general  instruction  in  "The  Manual 
for  Organization  and  Forest  Management  in  Siberian  Cedar 
Pine  Stands"  (Isaev  1990).  Now,  in  Siberia  great  experimen- 
tal works  are  being  conducted  on  a  comprehensive  estima- 
tion of  Siberian  stone  pine  forests,  on  their  selection  and 
growth,  and  on  the  creation  and  control  of  new  felling  tech- 
nology. The  practical  resolution  of  these  problems  will  per- 
mit the  rational  use  of  the  Siberian  stone  pine  resources,  the 
conservation  of  the  better  part  of  the  Siberian  stone  pine 
forests  for  ecological  purposes,  the  reproduction  of  valuable 
gene  soxu-ces,  and  the  enhancement  of  recreation  uses. 

Perhaps,  many  new  theoretical  principles  of  Siberian 
stone  pine  forest  management  may  be  of  use  with  other 
cedar  pines,  other  cedars,  or  other  conifers.  We  confirmed 
the  validity  of  many  of  our  suppositions  when  we  visited 
Korean  pine  forests  of  the  Great  Hingan  in  China. 

The  problem  of  reproducing  Siberian  stone  pine  forests 
was  believed  very  difficult  and  insoluble  for  a  long  time.  In 
the  1960's,  problems  with  seed  stratification  were  solved. 
At  first  results  were  not  positive,  but  now  the  technology 
of  seed  stratification  is  worked  out. 

The  problem  of  grovdng  Siberian  stone  pine  seedlings  has 
also  been  solved,  and  there  is  a  great  reserve  of  planting 
material  in  nurseries.  But  this  material  is  not  sold  because 
the  politics  of  reproducing  Siberian  stone  pine  forests  have 
changed  during  the  last  few  years.  At  first  the  main  method 
of  reproducing  Siberian  stone  pine  forests,  as  well  as  other 
coniferous  forests,  was  forest  plantations.  This  method  has 


319 


Table  8 — Intrapopulation  structure  of  crop  capacity  of  stand  (cones  per  tree) 


Statistical 
index 


Number 
of  cones 


Tree 


Diameter  Age 


Per 
tree 


Per 
shoot 


Number 

of 
female 
shoots 
per  tree 


Cvof 
crop  for 
10  years 


Number 
of  trees 


Number 
of  cones 


cm 


Years 


Per  ha  Percent 


Class  I:  0-0.50 


Per  ha  Percent 


X 

38.8 

186.5 

15.6 

1.06 

14.4 

50.8 

21 

30 

300 

8 

m 

1.2 

5.4 

1.2 

.04 

1.0 

1.8 

Cv 

22.7 

20.3 

54.6 

27.80 

49.0 

25.8 

Class  II:  0.51  - 1.00 

X 

51 .1 

21 1 .0 

AC  ^ 

45.7 

1 .37 

34.7 

38.4 

20 

30 

900 

22 

m 

1.1 

4.4 

1.3 

.05 

1.4 

1.5 

Cv 

15.8 

14.7 

19.3 

18.90 

28.3 

27.0 

Class  III:  1.01  - 1.50 

X 

55.7 

206.0 

72.3 

1 .41 

51 .6 

4 

37.1 

12 

17 

900 

oo 
22 

m 

2.2 

5.7 

2.1 

.04 

1.6 

1.5 

Cv 

21.2 

14.8 

15.6 

15.30 

16.3 

21.6 

Class  IV:  1.51  -  2.00 

X 

69.9 

223.7 

99.2 

1.36 

74.4 

38.6 

9 

13 

900 

22 

m 

3.9 

7.0 

1.8 

.04 

2.6 

1.7 

Cv 

27.5 

14.3 

8.4 

15.10 

15.8 

20.0 

Class  V:  more  2.01 

X 

62.9 

222.2 

150.6 

1.58 

96.5 

35.1 

7 

10 

1,000 

26 

m 

1.5 

5.8 

6.5 

.05 

5.2 

2.9 

Cv 

10.1 

10.8 

17.9 

14.70 

22.3 

34.4 

Mean 

X 

51.2 

205.5 

58.7 

1.31 

42.8 

41.6 

69 

100 

4,000 

100 

m 

1.1 

5.8 

3.4 

.02 

2.2 

.9 

Cv 

27.8 

17.1 

73.8 

23.40 

67.0 

29.7 

proven  unsuitable  dviring  30  years.  Forest  plantations  were 
created  by  planting  small  2-year-old  seedlings  in  harvested 
areas  that  had  become  overgrown  with  heavy  grass  cover 
and  intensive  growth  of  birch  and  aspen.  Because  the  for- 
est plantation  was  overgrown,  it  died.  The  necessary  care 
of  these  plantations  was  absent. 

Now,  the  main  method  for  reproducing  forests  is  by  con- 
serving imderstory  trees  in  Siberian  stone  pine  forests  during 
cleaning  operations.  Particular  hopes  in  Siberia  and  other 
regions  of  Russia  rest  on  an  initial  thinning  of  Siberian  stone 
pine  undergrowth  under  cover  of  birch  and  other  leaf-bearing 
trees.  Siberian  cedar  pine  normally  reproduces  naturally 
as  a  result  of  tree  succession  in  grassy  forest  types.  Our 
aim  is  to  assist  this  process. 

There  are  millions  of  hectares  of  natural  generation  of 
Siberian  stone  pine  under  birch  and  Scotch  pine  overstory. 
For  the  last  century,  at  least  in  Siberia,  the  intensive  macro- 
cycle  of  Siberian  stone  pine  generation  has  developed.  It 
includes  much  movement  of  the  Siberian  stone  pine  bound- 
ary area  in  the  north  and  in  the  mountains.  Theoretical 
principles  of  optimal  Siberian  stone  pine  forest  reproduction 
are  the  doctrine  for  reproduction  and  age  d)Tiamics  that 
Russian  scientists  developed  in  the  Ural  (Kolesnikov  and 
Smolonogov  1960;  Smolonogov  1990). 


Reproducing  Siberian  stone  pine  today  in  great  volimaes 
depends  mainly  on  correct  technology  of  felling  in  leaf- 
bearing  forests.  Now,  our  Institute  has  decided  to  accom- 
plish the  task  using  Finland  logging  machines. 

Artificial  reproduction  of  Siberian  stone  pine  is  restricted 
mainly  to  the  estabhshment  of  the  forest  plantations  for  dif- 
ferent purposes.  The  selection  forest  inventory,  the  selec- 
tion choice  of  plus  trees,  and  the  control  of  posterity  and 
reproduction  are  made  for  this.  The  selection  is  conducted 
for  growth  rate,  resin  productivity,  and  crop  capacity. 

SELECTING  CROP  FORMS 

The  choices  for  selecting  crop  forms  from  the  population 
are  great.  For  example,  the  data  for  intrapopulation  struc- 
ture of  crop  capacity  (table  8)  show  that  trees  of  V  class  were 
of  specific  interest.  They  exceed  the  crop  capacity  for  the  av- 
erage tree  in  the  stand  by  more  than  two  times.  These  trees 
have  high  cone-bearing  energy  (number  of  cones  per  shoot) 
and  well-developed  female  storey.  They  comprise  one-fourth 
of  the  stand  and  77  percent  of  the  crop.  This  is  more  than 
four  times  the  production  of  all  low-crop  trees.  The  better 
crop  capacity  trees  have  good  growth  indexes  too — their 
diameters  are  more  than  the  mean  diameters  of  sample 


320 


Table  9 — Characteristics  of  plus  trees  related  to  crop  capacity 


Number  of 


Number 
of  trees 

Tree 
Diameter 

Aae 

per 
tree 

Number  of  cones 
per  1  cm 
of 

diameter 

oer 
shoot 

cone- 
bearing 
shoots  per 
tree 

Cv  of 
crop  for 
1 0  years 

Selection 
rank 

cm 

Years 

17 

52 

198 

160 

3.1 

1.9 

83 

13.6 

2.8 

29 

64 

228 

154 

2.4 

1.9 

81 

29.0 

2.2 

30 

64 

180 

191 

2.9 

1.7 

115 

42.2 

2.6 

QC 
90 

Of 

1  /U 

l.O 

liiii 

o/.b 

98 

56 

190 

199 

3.4 

1.8 

107 

28.5 

3.1 

107 

60 

210 

163 

2.7 

1.6 

99 

31.8 

2.4 

128 

68 

190 

197 

2.8 

1.4 

142 

59.7 

2.5 

Mean  for 

class 

60 

199 

176 

2.8 

1.7 

108 

34.6 

2.5 

Mean  for 

stand 

51 

205 

58 

1.1 

1.3 

43 

41.6 

1.0 

trees  by  20  percent.  Examples  of  crop  capacity  plus  trees 
and  the  individual  variability  of  crop  structure  can  be  seen 
in  table  9. 

These  trees  are  nov5^  reproduced  in  selection  centers.  The 
basic  center  is  at  our  Institute  near  Tomsk.  Valuable  selec- 
tion plantations  are  created  by  planting  seedlings  of  50  cm 
length  with  close  root  systems  in  brickets  in  our  center.  We 
are  sure  that  this  technology  will  result  in  good  quahty  and 
conservation  of  Siberian  stone  pine  forest  plantations. 

Solution  of  these  problems  is  the  main  emphasis  of  work 
on  artificial  reproduction  of  Siberian  stone  pine  in  Siberia. 

REFERENCES 

Chaylakhyan,  M.  Kh.  1984.  Regulations  of  high  plant  flow- 
ering. In:  Hormonal  regulation  of  ontogenesis  in  plants. 
Nauka,  Moscow:  9-27.  [In  Russian]. 

Isaev,  A.  S.,  ed.  1990.  The  manual  for  organization  and  for- 
est management  in  Siberian  cedar  pine  stands.  Moscow. 
120  p.  [In  Russian]. 

Isaev,  A.  S.,  ed.  1985.  Siberian  cedar  pine  forests  of  Siberia. 
Nauka,  Novosibirsk.  256  p.  [In  Russian]. 

Kolesnikov,  B.  P.;  Smolonogov.  1960.  Some  principles  of 
age  and  regeneration  dynamics  in  Zaural'skoe  Priob'e 
cedar  pine  forests.  In:  Problems  of  cedar.  Novosibirsk: 
21-33.  [In  Russian]. 

Mosin,  V.  I.;  Savina,  L.  G.  1985.  Physiological  and  bio- 
chemical peculiarity  of  Scotch  Pine  for  different  sexual 
type.  In:  Sexual  reproduction  of  conifers.  Novosibirsk: 
59-60.  [In  Russian]. 


Rush,  V.  A.  1974.  Biochemical  characteristics  of  cedar  pine 
seeds.  In:  Biology  of  seed  reproduction  of  western  Siberian 
conifers.  Nauka,  Novosibirsk:  180-184.  [In  Russian]. 

Rush,  V.  A.;  Lizunova,  V.  V.  1969.  Some  principles  of  chemi- 
cal composition  of  Siberian  cedar  pine  seeds.  Vegetational 
Resources.  5(4):  519-525.  [In  Russian]. 

Samsonova,  A.  E.;  Bolgova,  T.  B.  1985.  Biochemical  aspects 
of  sexuahzation  of  Scotch  Pine  in  the  central  forest-steppe 
region.  In:  Sexual  reproduction  of  conifers.  Novosibirsk: 
60-62.  [In  Russian]. 

Smolonogov,  E.  P.  1990.  Ecological  and  geographical  differ- 
entiation and  dynamics  of  Siberian  cedar  pine  in  Urals 
and  West  Siberian  Plain.  Ural  Div.  USSR  Acad.  Sci.  Publ., 
Sverdlovsk.  286  p.  [In  Russian]. 

Vorobjeva,  N.  A.  1973.  Peculiarity  of  carbohydrate  metabo- 
lism in  Siberian  cedar  pine  from  activation  of  reproductive 
processes.  In:  Sexual  reproduction  of  conifers.  Nauka, 
Movosibirsk:  210-213.  [In  Russian]. 

Vorobjev,  V.  N.  1979.  A  method  of  retrospective  study  of 
seminiference  dynamics  in  Pinus  sihirica  Du  Toiir.  Bot. 
Zhurn.  64(7):  971-974.  [In  Russian]. 

Vorobjev,  V.  N.  1983.  Biological  principles  of  comprehen- 
sive utilization  in  Siberian  cedar  pine  forests.  Nauka, 
Novosibirsk.  256  p.  [In  Russian]. 

Vorobjev,  V.  N.;  Dubovenko,  Zh.  V.;  Pentegova,  V.  A.  1971. 
On  hemotaxonomic  characteristics  of  Pinus  sihirica  Du 
Tour  mountain  forms.  Vegetational  Resources.  7(4): 
564-567.  [In  Russian]. 

Vorobjev,  V.  N.;  Vorobjeva,  N.  A.;  Sviridenko,  E.  I.;  Kolesov, 
V.  M.  1979.  Siberian  cedar  pine  seeds.  Nauka,  Novosibirsk. 
129  p.  [In  Russian]. 


321 


Schmidt,  Wyman  C;  Holtmeier,  Friedrich-Karl,  comps.  1994.  Proceedings — international 
workshop  on  subalpine  stone  pines  and  their  environment:  the  status  of  our  knowl- 
edge; 1992  September  5-1 1 ;  St.  Moritz,  Switzerland.  Gen.  Tech.  Rep.  INT-GTR-309. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Service,  Intermountain  Research 
Station.  321  p. 

This  proceedings  is  the  product  of  the  first  international  workshop  on  the  five  subalpine 
stone  pines  of  the  world — Pinus  albicaulis,  P.  cembra,  P.  koraiensis,  P.  pumila,  and 
P.  sibirica.  It  includes  48  papers  on  the  evolution  and  taxonomy,  ecology,  regeneration, 
growth,  environmental  factors,  wildlife,  forest  structure  and  dynamics,  forest  manage- 
ment, and  research  needs  of  the  stone  pines.  All  five  stone  pines — one  in  North  America 
and  the  other  four  in  Europe  and  Asia — share  similar  flora  and  fauna  and  behave  in 
similar  ecological  patterns. 


KEYWORDS:  Pinus  albicaulis,  P.  cembra,  P.  koraiensis,  P.  pumila,  P.  sibirica,  wildlife, 
bears,  squirrels,  nutcracker  birds,  stone  pine  ecology,  regeneration,  seed 
dispersal,  taxonomy,  genetics,  growth,  research  needs 


Schmidt,  Wyman  C;  Holtmeier,  Friedrich-Karl,  comps.  1994.  Proceedings — international 
workshop  on  subalpine  stone  pines  and  their  environment:  the  status  of  our  knowledge; 
1992  September  5-11;  St.  Moritz,  Switzerland.  Gen.  Tech.  Rep.  INT-GTR-309.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Intermountain  Research  Station,  321  p. 

Dieser  Bericht  ist  das  Ergebnis  des  ersten  Internationalen  Workshops  uber  die  funf 
subalpinen  Steinkieferarten  der  Erde — Pinus  albicaulis,  P.  cembra,  P.  koraiensis,  P.  pumila 
und  Pinus  sibirica.  Er  entalt  48  Vortrage  uber  die  Evolution  und  Taxonomy,  die  Okologie, 
die  Regeneration,  das  Wachstum,  die  Standortfaktoren,  das  Tierleben,  die  Waldstruktur  und 
-dynamik,  die  Waldbewirtschaftung  und  -pflege  sowie  den  Forschungsbedarf.  Die  funf 
Steinkieferarten  -  eine  in  Nordamerika  und  vier  in  Eurasien  beheimatet  -  weisen  viele 
okologische  Gemeinsamkeiten  auf. 


STICHWORTE:  Pinus  albicaulis,  P.  cembra,  P.  koraiensis,  P.  pumila,  P.  sibirica,  Tierleben, 
Baren,  Eichhornchen,  Tannenhaher,  Okologie,  Regeneration, 
Samenverbreitung,  Taxonomy,  Genetik,  Wachstum,  Forschungsbedarf 


#1^ 

Federal  Recycling  Program        Printed  on  Recycled  Paper 


The  Intermountain  Research  Station  provides  scientific  knowledge  and  technology  to  im- 
prove management,  protection,  and  use  of  the  forests  and  rangelands  of  the  Intermountain 
West.  Research  is  designed  to  meet  the  needs  of  National  Forest  managers,  Federal  and 
State  agencies,  industry,  academic  institutions,  public  and  private  organizations,  and  individu- 
als. Results  of  research  are  made  available  through  publications,  symposia,  workshops, 
training  sessions,  and  personal  contacts. 

The  Intermountain  Research  Station  territory  includes  Montana,  Idaho,  Utah,  Nevada,  and 
western  Wyoming.  Eighty-five  percent  of  the  lands  in  the  Station  area,  about  231  million 
acres,  are  classified  as  forest  or  rangeland.  They  include  grasslands,  deserts,  shrublands, 
alpine  areas,  and  forests.  They  provide  fiber  for  forest  industries,  minerals  and  fossil  fuels  for 
energy  and  industrial  development,  water  for  domestic  and  industrial  consumption,  forage  for 
livestock  and  wildlife,  and  recreation  opportunities  for  millions  of  visitors. 

Several  Station  units  conduct  research  in  additional  western  States,  or  have  missions  that 
are  national  or  international  in  scope. 

Station  laboratories  are  located  in: 

Boise,  Idaho 

Bozeman,  Montana  (in  cooperation  with  Montana  State  University) 
Logan,  Utah  (in  cooperation  with  Utah  State  University) 
Missoula,  Montana  (in  cooperation  with  the  University  of  Montana) 
Moscow,  Idaho  (in  cooperation  with  the  University  of  Idaho) 
Ogden,  Utah 

Prove,  Utah  (in  cooperation  with  Brigham  Young  University) 

Reno,  Nevada  (in  cooperation  with  the  University  of  Nevada) 

The  policy  of  the  United  States  Department  of  Agriculture  Forest  Service  prohibits 
discrimination  on  the  basis  of  race,  color,  national  origin,  age,  religion,  sex,  or  disability, 
familial  status,  or  political  affiliation.  Persons  believing  they  have  been  discriminated 
against  in  any  Forest  Service  related  activity  should  write  to:  Chief,  Forest  Service,  USDA, 
P.O.  Box  96090,  Washington,  DC  20090-6090.